PRIVILEGED AND CONFIDENTIAL ATTORNEY WORK PRODUCT
CALDERWOOD & HARKNESS LLP 101 Montgomery Street, Suite 3500 San Francisco, California 94104
MEMORANDUM
FROM: Naomi Petersen, Working Associate, Mergers & Acquisitions, Calderwood & Harkness LLP
TO: Andrew Fitzpatrick, Lead Associate, Mergers & Acquisitions, Calderwood & Harkness LLP
CC: Harrison Wexford, Responsible Partner, Calderwood & Harkness LLP; Marcus Trejo, Partner, Head of Growth Equity, Pinnacle Capital Management LLC
DATE: July 12, 2022
RE: intellectual Property and Technology Due Diligence Summary — Verdana Software, Inc.
MATTER NO.: 1002-00004
CLIENT NO.: PIN-0042
SUBJECT: Summary of IP and technology diligence findings in connection with the proposed $60,000,000 Series D Preferred Stock investment by Pinnacle Venture Fund II LP ("Pinnacle" or the "Investor") in Verdana Software, Inc. ("Verdana" or the "Company")
I. EXECUTIVE SUMMARY
The purpose of this memorandum is to summarize the key findings from our intellectual property and technology due diligence review of Verdana Software, Inc. in connection with the proposed $60,000,000 minority growth equity investment by Pinnacle Venture Fund II LP. Verdana is a Delaware C-corporation (EIN 93-1485726; Delaware Secretary of State File Number 6274913) headquartered at 1175 NW Couch Street, Suite 800, Portland, Oregon 97209. The Company was founded in 2016 by Raj Anand (Chief Executive Officer & Co-Founder) and Lena Kovac (Chief Technology Officer & Co-Founder), and as of the date of this diligence review employs approximately 310 individuals.
Our diligence commenced on May 16, 2022, following execution of the non-disclosure agreement between Pinnacle and Verdana. Access to the Company's virtual data room was provided by Langford Price & Associates LLP, Verdana's outside corporate counsel, through its representatives Jonathan Hale, Partner, and Rebecca Thornton, Senior Associate. The data room was hosted on Thackeray Solutions and organized according to the diligence request list dated May 14, 2022, which we prepared in coordination with the Pinnacle deal team.
Verdana's core product is the Verdana SCX platform, an artificial intelligence-powered supply chain optimization software-as-a-service ("SaaS") platform sold primarily to mid-market and enterprise manufacturers and logistics companies under multi-year subscription contracts. The platform leverages proprietary machine learning algorithms—including deep learning and reinforcement learning models—to deliver demand forecasting, inventory optimization, logistics routing, supplier risk scoring, and real-time analytics capabilities to its customer base. As of Q1 2022, the Company reported annual recurring revenue ("ARR") of approximately $47.3 million, representing year-over-year growth of approximately 63.7%.
Overview of Findings. Verdana maintains a material and growing intellectual property portfolio centered on its proprietary AI/ML algorithms and the SCX platform. The Company's IP ownership and assignment practices are generally well-organized, and its patent, trademark, and trade secret portfolios are consistent with what we would expect from a growth-stage enterprise SaaS company at this valuation level. However, our review has identified certain areas that warrant follow-up and, in some cases, remediation in connection with the proposed transaction, including: (i) gaps in employee and contractor intellectual property assignment documentation, particularly among early hires and founding-era contractors; (ii) open-source software license compliance concerns, specifically with respect to one AGPL v3-licensed component incorporated into a customer-facing backend microservice; and (iii) a pending third-party patent infringement allegation from Optivex Analytics, Inc. that, while preliminary and unlitigated, merits monitoring and appropriate transactional protections.
Overall Risk Rating: Moderate. No issues have been identified that we believe would be transaction-blocking for the proposed Series D investment. However, we recommend that certain remediation steps be addressed either as conditions to closing or as post-closing covenants, as detailed in Section XIV of this memorandum. The flagged items are consistent in nature and scope with issues commonly encountered in growth-stage technology companies, and each can be addressed through standard transactional mechanisms.
II. SCOPE OF REVIEW
Our review encompassed the following categories of documents and information sources:
Documents and Materials Reviewed:
• Organizational intellectual property records, including the Company's internal IP register and asset inventory
• Patent filings and prosecution files for all issued U.S. patents and pending U.S. and international patent applications
• Trademark registrations, applications, and related prosecution correspondence for all U.S. and international trademarks
• Copyright registrations (to the extent applicable)
• Domain name portfolio and registration records
• Trade secret policies, procedures, and the Company's trade secret register
• Software development documentation, including architecture overviews, technology stack descriptions, and development process documentation
• Open-source software ("OSS") audit reports, including the independent audit conducted in Q4 2021
• Internal OSS inventory/registry maintained by the engineering team
• Employee Confidential Information and Invention Assignment Agreements ("CIIAs") and related HR records
• Independent contractor agreements, including IP assignment and work-for-hire provisions
• Third-party license agreements (inbound), including cloud infrastructure, data, and software library licenses
• Customer agreements (outbound) and related IP provisions, with particular focus on the top five customer relationships by ARR
• Correspondence regarding any IP disputes, infringement claims, or cease-and-desist letters, including the Optivex Analytics, Inc. matter
• Data room indices and documents contained in Section 4 (Intellectual Property), Section 5 (Technology), and Section 8 (Litigation) of the virtual data room, organized pursuant to the diligence request list dated May 14, 2022
• Verdana's most recent SOC 2 Type II audit report (Q4 2021)
• Co-founder IP assignment agreements executed at incorporation
Interviews and Management Discussions:
• Diligence call with Lena Kovac, Chief Technology Officer & Co-Founder, conducted on June 20, 2022, covering the SCX platform architecture, development practices, OSS usage, and the AGPL component remediation plan
• Diligence call with Sandra Mikkelsen, General Counsel, conducted on June 22, 2022, covering IP ownership and assignment practices, the Optivex Analytics claim, trade secret protections, and the status of missing employee CIIAs
• Follow-up written Q&A responses received from Company counsel (Langford Price & Associates LLP) on June 28, 2022, addressing supplemental questions arising from the June 20 and June 22 calls
In total, we reviewed approximately 340 documents across the three data room sections and related supplemental materials. Where referenced in this memorandum, specific documents are identified by their data room index numbers for ease of cross-reference.
III. VERDANA SCX PLATFORM — TECHNOLOGY OVERVIEW
A. Platform Description
The Verdana SCX platform is a cloud-native, multi-tenant business-to-business SaaS platform designed to provide AI-driven supply chain optimization capabilities to mid-market and enterprise customers. The platform's name derives from "Supply Chain Xcelerate," which also serves as the Company's registered tagline. The SCX platform is the Company's sole commercial product and the source of substantially all of its revenue.
Core Functionality. The SCX platform provides the following core functional modules to its customers:
• Demand Forecasting: Utilizes proprietary deep neural network models to generate granular demand forecasts across product lines, geographies, and time horizons, enabling customers to anticipate demand shifts and adjust procurement and production planning accordingly.
• Inventory Optimization: Employs reinforcement learning algorithms to dynamically optimize inventory levels across multiple warehouses and distribution centers, balancing carrying costs against service level targets and lead time variability.
• Logistics Routing: Provides multi-modal logistics routing optimization, incorporating real-time data feeds (e.g., weather, traffic, port congestion) to recommend optimal shipping routes and carrier selections.
• Supplier Risk Scoring: Leverages a proprietary graph-based scoring system to assess and monitor supplier risk across financial, operational, geopolitical, and environmental dimensions, drawing on both customer-provided data and third-party data sources.
• Real-Time Analytics Dashboards: Delivers configurable dashboards and reporting tools that aggregate supply chain performance metrics, alert users to anomalies, and enable scenario modeling.
Technology Stack. The SCX platform is built primarily on a Python and Go microservices architecture, which provides modular scalability and facilitates independent deployment of platform components. The platform is hosted on Ridgeway Industries ("Meridian Technologies") under a standard enterprise cloud infrastructure services agreement. Data storage and management rely on PostgreSQL for structured relational data and a proprietary graph database purpose-built by the Verdana engineering team for supply chain network modeling. The proprietary graph database is a key technical differentiator and is maintained as a trade secret (discussed further in Section VI below).
Key Technical Differentiator. The Company's primary competitive advantage resides in its proprietary machine learning models, which include both deep learning architectures for demand forecasting and reinforcement learning frameworks for inventory and logistics optimization. These models are trained on aggregated, anonymized customer supply chain data accumulated over the platform's operational history. According to Lena Kovac, the depth and breadth of this training data set — which spans multiple industries, geographies, and supply chain configurations — represents a significant and growing competitive moat, as the models' predictive accuracy improves with each additional customer deployment and data contribution. The Company considers these models, and the associated training data sets, to be among its most valuable trade secrets.
Financial Metrics. As of Q1 2022, the Company reported the following key performance indicators relevant to the technology diligence assessment:
• Annual Recurring Revenue (ARR): $47.3 million (up from $28.9 million in Q1 2021, representing approximately 63.7% year-over-year growth)
• Gross Margin: 74.2%
• Net Revenue Retention (NRR): 128%
These metrics indicate strong product-market fit and meaningful expansion revenue from existing customers, which is consistent with a platform that delivers measurable value and retains customer engagement through proprietary technology.
Customer Base. The Company serves approximately 185 customers across the manufacturing, logistics, distribution, and related industries. Revenue concentration is notable, with the top five customers representing a meaningful share of total ARR:
Customer concentration risk, particularly with respect to the Meridian Logistics Corp. relationship, is addressed in the Customer Concentration Risk Analysis Memorandum dated July 5, 2022, and is not the primary focus of this IP and technology diligence review. However, as discussed in Section VI below, the significance of Meridian's data contributions to the platform's training data sets warrants confirmation that the Company's data aggregation practices are fully compliant with the terms of the Meridian customer agreement.
B. Development Team and Practices
The Company's engineering team comprises approximately 120 employees, representing roughly 39% of the Company's total headcount of approximately 310. The engineering organization is led by Lena Kovac, who serves as both CTO and Co-Founder, and is organized into functional teams covering platform engineering, machine learning and data science, infrastructure and DevOps, quality assurance, and product design.
Development Methodology. The engineering team follows an Agile/Scrum development methodology with two-week sprint cycles. The Company maintains a continuous integration/continuous deployment ("CI/CD") pipeline that enables frequent, incremental releases to the production SaaS environment. According to Lena Kovac, the team deploys to production approximately 15–20 times per week, with automated testing gates and staged rollout procedures to mitigate deployment risk.
Source Code Management. All source code is managed in an internally hosted Git-based version control system (Aldermere Software Enterprise Edition, self-hosted on the Company's Meridian Technologies infrastructure). The Company maintains structured code review processes requiring peer review and approval before code is merged into production branches. Access to the source code repository is controlled through role-based access controls, with access tiered based on team membership and seniority. These practices are consistent with industry standards and contribute to the Company's trade secret protection posture (as discussed in Section VI).
Outsourced Development. Based on our review of contractor agreements and management representations, the Company does not have material reliance on outsourced offshore development for its core platform. Limited use of independent contractors has been identified, primarily for specialized UI/UX design work and certain data science consulting engagements. As discussed in Section VII below, certain early-stage contractor agreements present IP assignment concerns that should be addressed in connection with the transaction.
IV. PATENT PORTFOLIO
A. Issued Patents
Verdana holds three (3) issued United States utility patents, each of which is relevant to the core technology underlying the SCX platform. The details of these patents are as follows:
1. U.S. Patent No. 10,671,284 — "System and Method for Real-Time Supply Chain Demand Forecasting Using Deep Neural Networks"
• Filing Date: August 14, 2018
• Issue Date: June 2, 2020
• Assignee: Verdana Software, Inc.
• Named Inventors: Lena Kovac, David Chen, Priya Ramanathan
• Status: Active; maintenance fees current through the next maintenance window
This patent covers the Company's core demand forecasting methodology, including specific claims directed to the architecture of the deep neural network used for time-series demand prediction in supply chain contexts, the data preprocessing pipeline for heterogeneous supply chain data inputs, and the real-time inference system that delivers forecasts to end users via the SaaS platform. The patent contains 22 claims (4 independent claims and 18 dependent claims).
2. U.S. Patent No. 11,042,597 — "Automated Inventory Optimization Through Reinforcement Learning"
• Filing Date: March 22, 2019
• Issue Date: June 22, 2021
• Assignee: Verdana Software, Inc.
• Named Inventors: Lena Kovac, Marco Ferretti
• Status: Active; maintenance fees current
This patent covers the Company's reinforcement learning-based approach to dynamic inventory optimization across distributed warehouse networks. The patent includes claims directed to the reward function design, the multi-agent reinforcement learning framework, and the integration of real-time supply chain event data into the optimization loop. The patent contains 18 claims (3 independent claims and 15 dependent claims).
3. U.S. Patent No. 11,138,215 — "Graph-Based Supplier Risk Scoring System"
• Filing Date: July 9, 2019
• Issue Date: October 5, 2021
• Assignee: Verdana Software, Inc.
• Named Inventors: Lena Kovac, Amara Osei, Jason Whitfield
• Status: Active; maintenance fees current
This patent covers the Company's proprietary graph-based methodology for assessing and scoring supplier risk, including specific claims directed to the graph data structure used to model multi-tier supplier relationships, the algorithm for propagating risk signals through the graph, and the dynamic updating mechanism that incorporates real-time data feeds. The patent contains 16 claims (3 independent claims and 13 dependent claims).
Assignment Status. We reviewed executed patent assignment agreements for each of the three issued patents. In each case, all named inventors have executed assignments in favor of Verdana Software, Inc., and the assignments have been recorded with the United States Patent and Trademark Office ("USPTO"). The assignment chains appear clean and unbroken. We did not identify any outstanding obligations, reversionary rights, or encumbrances with respect to any of the issued patents, other than the general security interest granted to Pacific Summit Bank, N.A. under the Company's revolving credit facility (discussed in Section X below).
Maintenance Fees. All maintenance fees for the three issued patents are current as of the date of this memorandum. The next maintenance fee due date for the earliest-issued patent (U.S. Patent No. 10,671,284) is June 2, 2024 (the 4-year maintenance fee window). Prosecution and maintenance are managed by Breckenridge IP Group LLP, a separate IP boutique firm retained by Verdana for patent matters.
B. Pending Patent Applications
Verdana has three (3) pending patent applications, consisting of two U.S. utility patent applications and one international application under the Patent Cooperation Treaty ("PCT"):
1. U.S. Patent Application No. 17/348,621 — "Multi-Modal Logistics Routing Optimization Engine"
• Filing Date: June 15, 2021
• Status: Office action received from the USPTO examiner; response due within the standard response period. Breckenridge IP Group LLP is preparing the response.
• Applicant: Verdana Software, Inc.
• Named Inventors: Lena Kovac, Thomas Bergström, Kenji Nakamura
The office action raised obviousness rejections (35 U.S.C. § 103) with respect to certain dependent claims. According to Breckenridge IP Group, the independent claims were not rejected and the office action is not considered a significant prosecution setback. A response narrowing the dependent claims is expected to be filed within the response window.
2. U.S. Patent Application No. 17/589,304 — "Federated Learning Framework for Cross-Enterprise Supply Chain Data"
• Filing Date: January 31, 2022
• Status: Pending; awaiting first office action from the USPTO. No substantive prosecution issues have arisen to date.
• Applicant: Verdana Software, Inc.
• Named Inventors: Lena Kovac, David Chen, Sophie Laurent
This application relates to a next-generation capability that would enable federated learning across multiple enterprise customers' supply chain data sets without requiring centralized data aggregation, which has potential implications for data privacy and competitive sensitivity. Lena Kovac indicated during the June 20 diligence call that this technology is currently in development and is expected to be commercially deployed in 2023.
3. PCT International Application No. PCT/US2021/036894 — Corresponding to U.S. Patent No. 10,671,284 ("System and Method for Real-Time Supply Chain Demand Forecasting Using Deep Neural Networks")
• Filing Date: June 11, 2021 (claiming priority to the U.S. patent)
• Designated Countries/Regions: European Union, United Kingdom, Japan, South Korea
• Status: National phase entry deadlines are being tracked by Breckenridge IP Group LLP; no substantive prosecution issues reported
We reviewed the prosecution files for each pending application and did not identify material prosecution issues beyond the routine office action described above. The applications are being managed by Breckenridge IP Group LLP, and the Company appears to be actively prosecuting its patent portfolio in a manner consistent with its commercial strategy and geographic expansion plans.
C. Freedom-to-Operate Considerations
Based on our review and discussions with Company management, Verdana has not conducted a formal, comprehensive freedom-to-operate ("FTO") analysis for the SCX platform. While the Company's outside patent counsel at Breckenridge IP Group has conducted prior art searches in connection with the Company's own patent prosecution, these searches were directed at patentability rather than freedom-to-operate, and do not constitute an FTO clearance.
The absence of a formal FTO analysis is not unusual for a company at Verdana's stage and is consistent with what we commonly observe in growth-stage enterprise SaaS companies. However, given the increasingly competitive landscape in AI-powered supply chain optimization — a field in which multiple competitors and established enterprise software incumbents hold patent portfolios — the lack of a comprehensive FTO study represents a risk factor that Pinnacle should consider.
Recommendation: Pinnacle should consider requesting that Verdana commission a formal FTO study as a post-closing covenant or as a diligence supplement to be completed prior to closing, with a particular focus on the demand forecasting, inventory optimization, and logistics routing modules of the SCX platform. The relevance of this recommendation is underscored by the pending Optivex Analytics claim discussed in Section IX below, which, while preliminary, highlights the potential for third-party patent assertions in this technology space.
V. TRADEMARK PORTFOLIO
A. Registered Trademarks
Verdana maintains the following registered trademarks in the United States:
All U.S. trademark registrations are current, and the Company has timely filed all required maintenance documents, including Sections 8 and 15 declarations under the Lanham Act where applicable. The trademark portfolio is managed by Langford Price & Associates LLP as part of the Company's general corporate counsel relationship.
B. Pending Trademark Applications
The Company has one (1) pending U.S. trademark application:
• VERDANA INSIGHTS — U.S. Trademark Application Serial No. 97/283,641, filed March 8, 2022, for Class 42 (SaaS services, specifically data analytics and business intelligence tools). The application is currently awaiting examination by the USPTO. No office actions or third-party oppositions have been received as of the date of this memorandum.
According to Sandra Mikkelsen, the VERDANA INSIGHTS mark is associated with a planned analytics and business intelligence add-on module for the SCX platform that the Company intends to launch commercially in late 2022 or early 2023.
International Trademark Registrations. The Company has obtained trademark registrations for the VERDANA SCX mark in the following jurisdictions:
These international registrations are consistent with the Company's current and anticipated customer base in North America and Europe, and align with the geographic designations of the pending PCT patent application.
C. Domain Names
Verdana holds the following key domain names, all of which are currently registered and in good standing:
• verdanascx.com (primary product domain)
• verdanasoftware.com (corporate domain)
• verdanascx.io
• verdanascx.co
• verdanasoftware.io
• verdanainsights.com (reserved for planned product launch)
Domain registrations are managed through a standard enterprise domain registrar. No domain name disputes, including UDRP (Uniform Domain-Name Dispute-Resolution Policy) proceedings or cybersquatting complaints, have been identified.
D. Observations. The Company's trademark portfolio is well-maintained and reflects consistent usage across its brand, product, and marketing materials. The marks are actively used in commerce in a manner consistent with the registrations, and the Company appears to exercise appropriate brand governance.
No opposition proceedings, cancellation actions, or other adverse trademark proceedings are pending with respect to any of Verdana's trademarks, either domestically or internationally.
We note, however, that the "VERDANA" word mark may face potential confusion arguments in certain international jurisdictions, particularly with respect to other software-related marks that bear phonetic or visual similarity. Sandra Mikkelsen indicated during the June 22 diligence call that the Company is aware of this potential issue and has monitored the relevant trademark registers but has not, as of the date of this memorandum, received any formal challenges, opposition notices, or infringement allegations related to the VERDANA mark. We recommend continued monitoring, particularly if the Company expands its trademark registrations into additional international jurisdictions.
VI. TRADE SECRETS AND PROPRIETARY KNOW-HOW
Verdana's most commercially significant intellectual property assets are its proprietary AI/ML algorithms, the associated training data sets derived from aggregated and anonymized customer supply chain data, the proprietary graph database architecture used for supply chain network modeling, and the broader engineering methodologies and know-how that underpin the SCX platform's competitive differentiation. These assets are protected primarily through trade secret law rather than (or in addition to) patent protection, and they constitute a substantial portion of the Company's enterprise value.
Trade Secret Protection Measures. Based on our review and management discussions, the Company maintains the following trade secret protection practices:
(a) Employee Confidentiality and Invention Assignment Agreements (CIIAs). All employees are required to sign CIIAs at the time of hire, which include confidentiality obligations covering trade secrets and proprietary information, as well as assignment provisions for inventions and works created during the course of employment. The gaps identified in CIIA coverage are discussed in Section VII below.
(b) Role-Based Access Controls. Access to the Company's source code repository, proprietary algorithms, and training data sets is restricted through role-based access controls within the Company's Aldermere Software and Meridian Technologies environments. Not all employees have access to all code; access is tiered based on team membership, job function, and seniority. Lena Kovac confirmed that access to the most sensitive ML model code and training data pipelines is limited to members of the machine learning and data science team (approximately 25 individuals).
(c) Visitor and Vendor NDAs. The Company requires non-disclosure agreements from all visitors to its offices and from all third-party vendors and partners who may have access to proprietary information.
(d) Data Room Access Logs. The Company maintains access logs for its internal systems and for external data sharing environments, including the diligence data room.
Trade Secret Register. Verdana maintains an internal trade secret register that identifies the Company's key proprietary algorithms, data assets, and technical methodologies. The register was last updated in Q1 2022 and includes, among other items, the demand forecasting neural network architecture, the reinforcement learning inventory optimization framework, the graph database schema and supplier risk propagation algorithm, and the aggregated training data sets. The existence of a formal trade secret register is a positive indicator of the Company's awareness of and commitment to trade secret protection, and we consider this practice to be above average for a company at Verdana's stage.
Aggregated Customer Data. A critical component of Verdana's trade secret portfolio is the aggregated, anonymized customer supply chain data that serves as the training corpus for the Company's machine learning models. We reviewed the IP and data-related provisions of the Company's standard customer agreement and representative agreements with the top five customers by ARR. Each reviewed agreement contains provisions that permit Verdana to use customer data in anonymized and aggregated form for purposes of product improvement, model training, benchmarking, and related analytics, subject to appropriate anonymization and de-identification requirements.
Recommendation: Given the significance of the aggregated training data to the Company's competitive moat and the concentration of data contributions from the Company's largest customers, we recommend that the Company confirm — through a specific review coordinated with Langford Price & Associates — that its data aggregation and anonymization practices are fully compliant with the terms of each of its top five customer agreements, with particular attention to the Meridian Logistics Corp. agreement (which accounts for 24.9% of ARR and is likely a proportionally significant contributor to the training data corpus). Any deviation from agreed-upon data handling practices could expose the Company to breach of contract claims or, in a worst-case scenario, to loss of the right to use the affected data.
VII. EMPLOYEE AND CONTRACTOR IP ASSIGNMENT AGREEMENTS
A. Employee Agreements
The Company uses a standard form Confidential Information and Invention Assignment Agreement ("CIIA") across its employee base. We reviewed the template CIIA and found it to be generally consistent with market-standard terms for technology companies, covering: (i) confidentiality obligations regarding proprietary information; (ii) present assignment of all inventions, works of authorship, and other intellectual property created by the employee in the course of employment or using Company resources; (iii) disclosure obligations for inventions; (iv) cooperation obligations for IP prosecution and enforcement; and (v) post-termination obligations regarding return of proprietary materials and continued confidentiality.
The CIIA template also contains the notice required under Oregon's employee invention statute (ORS 653.295), which provides that an employer may not require an employee to assign inventions that the employee develops entirely on the employee's own time without use of the employer's equipment, supplies, facilities, or trade secret information, unless the invention relates to the employer's business or actual or demonstrably anticipated research or development. Inclusion of this statutory notice is required under Oregon law and its presence in the template is appropriate given that Oregon law governs the Company's employment relationships.
Coverage Assessment. We reviewed HR records provided by the Company and cross-referenced them against the current employee roster. CIIAs are on file for approximately 295 of the Company's approximately 310 current employees. This represents coverage of approximately 95.2% of the current workforce.
Gap Identified. Approximately 15 employees do not have executed CIIAs on file. These employees are primarily early hires from the 2016–2017 period, during which the Company was in its founding and initial development phase and its HR and legal infrastructure was less formalized. Notably, some of these early hires occupy or previously occupied engineering roles, which increases the relevance of the gap given the possibility that these individuals contributed to foundational platform code or algorithms during the Company's early development period.
Sandra Mikkelsen indicated during the June 22 diligence call that the Company is aware of this gap and has initiated a process to obtain executed CIIAs from each of these individuals. As of the date of this memorandum, the Company has obtained executed agreements from 4 of the 15 identified employees, with the remaining 11 in process. Ms. Mikkelsen expressed confidence that the Company would be able to obtain the remaining agreements on a voluntary basis without material difficulty, but acknowledged that certain long-tenured employees may request review by personal counsel, which could introduce delays.
Founders. We reviewed the IP assignment agreements executed by each of the two co-founders, Raj Anand and Lena Kovac, at the time of the Company's incorporation in 2016. Each agreement provides for the assignment of all intellectual property related to the Company's business that was conceived, developed, or reduced to practice by the respective founder, both prior to and during the course of employment with the Company. The agreements are executed, appear to be in proper form, and cover the relevant subject matter. We did not identify any carve-outs, reversionary rights, or other provisions that would compromise the completeness of the founder IP assignments.
B. Contractor Agreements
We reviewed independent contractor agreements for approximately 25 contractors engaged by the Company since 2019. These agreements uniformly contain work-for-hire provisions designating the Company as the author of all works of authorship created by the contractor in the course of the engagement, as well as present-tense IP assignment clauses assigning all inventions, discoveries, improvements, and other intellectual property developed by the contractor in the course of the engagement to the Company. The agreements reviewed are consistent with market-standard terms and do not present material IP ownership concerns.
Gap Identified. Three (3) early-stage contractors, engaged during the 2016–2017 period for initial platform development work, have agreements with ambiguous IP assignment language. Specifically, these agreements assign "work product" to the Company but do not explicitly address the assignment of pre-existing intellectual property or background IP that may have been incorporated by the contractor into deliverables. Under the terms of these agreements, it is possible — though not certain — that certain pre-existing IP or tools brought by these contractors into the engagement remain owned by the contractors, even though such IP may now be embedded in or inseparable from the Company's codebase.
Based on our discussions with Lena Kovac and our review of the Company's development history, the SCX platform has undergone substantial evolution since the 2016–2017 period. Significant portions of the original codebase have been refactored or replaced as the platform matured. However, certain foundational architectural components — particularly elements of the data ingestion pipeline and the initial graph database schema — may trace their origins to work performed by these early contractors. Accordingly, while the risk is assessed as low-to-moderate, it cannot be entirely dismissed without further technical analysis or, preferably, confirmatory assignments.
Recommendation: Verdana should obtain confirmatory IP assignments from each of these three contractors as a condition to closing or as a post-closing covenant with a specified deadline. The confirmatory assignments should explicitly cover all intellectual property, including pre-existing and background IP, that was conceived, developed, or incorporated into deliverables for the Company during the course of each contractor's engagement. We note that obtaining such confirmatory assignments after the fact may require the Company to provide nominal additional consideration to the contractors. If any of the three contractors is uncooperative or cannot be located, the Company and its counsel should assess the technical scope of the affected contributions and evaluate whether a "clean room" reimplementation of any at-risk components is warranted.
VIII. OPEN-SOURCE SOFTWARE USAGE
A. Overview. The Verdana SCX platform incorporates various open-source software ("OSS") components, which is standard and expected practice in modern SaaS software development. The use of open-source components enables faster development cycles, leverages community-maintained code for non-differentiating functionality, and reduces development costs. However, OSS usage carries compliance obligations that vary depending on the applicable open-source license terms, and failure to comply with certain copyleft-style licenses can, in theory, expose proprietary code to source code disclosure obligations.
The Company's engineering team maintains an internal OSS inventory and registry, which catalogues the open-source components in use across the SCX platform, including the applicable license for each component, the version in use, and the manner in which the component is integrated into the platform (e.g., dynamically linked, statically linked, standalone tool). Lena Kovac indicated during the June 20 diligence call that the engineering team reviews new OSS additions prior to incorporation into the codebase, although this review process is informal and not governed by a formal written policy (as discussed in subsection C below).
In Q4 2021, the Company commissioned an independent OSS audit, the report of which was provided in the data room and reviewed by our team. The audit identified approximately 180 distinct OSS components in use across the SCX platform codebase. The audit was conducted by an independent software composition analysis firm and included both automated scanning and manual review of the Company's code repositories.
B. License Categories
The OSS components identified in the audit and internal registry fall into the following license categories:
Permissive Licenses (No Material Compliance Concerns). The majority of OSS components used by the Company are licensed under permissive open-source licenses, including the MIT License, the Apache License 2.0, and various BSD licenses (2-Clause and 3-Clause). Permissive licenses generally require only attribution and reproduction of the license text in distributed software, and do not impose copyleft or source code disclosure obligations. We did not identify material compliance concerns with respect to the Company's use of permissively licensed components.
Copyleft Licenses (Requires Further Analysis). Approximately 12 OSS components were identified as being licensed under copyleft licenses, including the GNU General Public License v2 ("GPL v2"), the GNU General Public License v3 ("GPL v3"), the GNU Lesser General Public License ("LGPL"), and the GNU Affero General Public License v3 ("AGPL v3"). Copyleft licenses require, in varying degrees and under varying conditions, that derivative works or works that incorporate copyleft-licensed code be distributed under the same license terms, which can include an obligation to make source code available to recipients. The analysis of these 12 components is as follows:
(i) LGPL Components (8 components) — Low Risk. Eight of the twelve copyleft-licensed components are licensed under the LGPL (versions 2.1 and 3.0). These components are used as standalone libraries that are dynamically linked to the Company's proprietary code. The LGPL generally permits dynamic linking to proprietary software without triggering copyleft obligations for the proprietary code, provided that the LGPL-licensed library itself remains modifiable and replaceable by end users. Because the SCX platform is delivered as a hosted SaaS product (rather than distributed software), the traditional LGPL dynamic linking analysis is somewhat attenuated — end users do not receive copies of the software. Accordingly, we assess the risk associated with these eight LGPL components as low.
(ii) GPL v3 Components Used in Internal Tooling (2 components) — Minimal Risk. Two components licensed under GPL v3 are used exclusively in the Company's internal build, test, and development tooling environments. These components are not distributed to customers, are not included in the SaaS product delivered to end users, and are not made available over a network in a manner that would trigger any distribution or network-use obligations. The risk associated with these components is assessed as minimal.
(iii) AGPL v3 Component in Backend microservice (1 component) — Moderate Risk; Flagged. One utility library licensed under the AGPL v3 has been incorporated into a backend microservice that interfaces with customer-facing functionality of the SCX platform. The AGPL v3 is notable among open-source licenses for its "network use" provision (Section 13 of the AGPL v3), which extends the copyleft obligation to situations where the software is made available for interaction over a computer network, even without traditional "distribution" of copies to users. If the Company's incorporation of this AGPL-licensed library into its backend microservice constitutes making the AGPL-licensed software available to users over a network, the Company may be obligated to make the complete source code of the "corresponding source" available to those users — which could, depending on the interpretation of "corresponding source" and the degree of integration, encompass the Company's proprietary microservice code.
This is the most significant OSS compliance issue identified in our review. Lena Kovac indicated during the June 20 diligence call that the engineering team is aware of this issue and is actively evaluating replacement of the AGPL-licensed utility library with a functionally equivalent alternative available under a permissive open-source license (MIT or Apache 2.0). Ms. Kovac estimated that the replacement could be completed by the end of Q3 2022, subject to testing and validation. We consider this remediation timeline to be reasonable and achievable.
(iv) GPL v3 Component with Potential Static Linking (1 component) — Low-to-Moderate Risk; Flagged. One data processing library licensed under GPL v3 may be statically linked in one module of the platform. Static linking of GPL v3-licensed code into proprietary code is generally understood to create a "derivative work" under the GPL, which would trigger the copyleft obligation for the entire combined work. However, the Company's engineering team is conducting further technical analysis to determine whether the linkage in question is truly static or whether the module can be refactored to use dynamic linking or an alternative approach. This analysis was ongoing as of the date of this memorandum.
C. OSS Compliance Program
The Company does not currently maintain a formal, written open-source software compliance policy. While informal practices exist — including engineering-level review of new OSS additions before incorporation, maintenance of the internal OSS registry, and the commissioning of the Q4 2021 independent audit — these practices have not been codified into a documented policy with defined approval workflows, escalation procedures, or regular audit schedules.
The absence of a formal OSS compliance policy is a gap, though one that is not uncommon among growth-stage technology companies. As the Company scales and as its codebase and contractor/contributor base grow, the risk of inadvertent OSS non-compliance increases, making a formal policy increasingly important.
Recommendations:
(i) Pinnacle should request, as a post-closing covenant in the transaction documents, that Verdana adopt a formal, written OSS compliance policy within a specified period following closing (e.g., 90 days). The policy should address, at minimum: (a) a classification of permissible and restricted OSS license categories; (b) a mandatory approval workflow for the introduction of new OSS components into the codebase; (c) periodic (at least annual) automated and/or manual OSS audits; (d) designated responsibility for OSS compliance oversight within the engineering or legal organization; and (e) remediation procedures for identified compliance issues.
(ii) The AGPL v3 component described above should be remediated — i.e., replaced with a permissive-licensed alternative — either prior to closing or promptly following closing pursuant to a post-closing covenant with a hard deadline (we suggest no later than September 30, 2022, consistent with the engineering team's estimated Q3 2022 completion timeline).
(iii) The Company should complete its technical analysis of the GPL v3 static linking issue and, if static linking is confirmed, remediate the affected module by refactoring to eliminate the static linkage or by replacing the GPL-licensed component.
IX. THIRD-PARTY IP CLAIMS AND DISPUTES
A. Pending Claim — Optivex Analytics, Inc.
In March 2022, Verdana received a cease-and-desist letter from Optivex Analytics, Inc. ("Optivex"), a Delaware corporation headquartered in Austin, Texas. The letter alleges that certain features of the Verdana SCX platform's demand forecasting module infringe U.S. Patent No. 10,489,137, owned by Optivex and titled "Predictive Analytics Methods for Supply Chain Event Detection." The Optivex patent was issued in November 2019 and contains claims directed to methods of using predictive analytics to detect and classify anomalous events in supply chain data streams. The patent lists three named inventors, none of whom are or have been affiliated with Verdana.
The cease-and-desist letter was addressed to Sandra Mikkelsen, General Counsel, and asserted that Verdana's demand forecasting module practices at least claims 1, 4, 7, and 12 of the Optivex patent. The letter demanded that Verdana immediately cease all allegedly infringing activities and enter into licensing discussions with Optivex. The letter did not include detailed claim charts or technical analysis supporting the infringement allegations.
Company Response. Sandra Mikkelsen responded to the Optivex letter in April 2022, denying that the Verdana SCX platform infringes any valid claim of the Optivex patent and requesting that Optivex provide detailed claim charts and a technical analysis mapping the asserted patent claims to specific features of the Verdana platform. As of the date of this memorandum, Optivex has not responded to the Company's request, and no further correspondence has been exchanged. No litigation has been filed by Optivex, and no litigation hold or preservation notice has been issued by either party.
Preliminary Non-Infringement Analysis. Following receipt of the Optivex letter, the Company's outside patent counsel at Breckenridge IP Group LLP conducted a preliminary non-infringement analysis. Based on our review of the draft analysis, Breckenridge IP Group concluded that the Verdana SCX demand forecasting approach differs materially from the methods claimed in the Optivex patent in several respects, including: (i) the Verdana system uses deep neural network architectures for demand prediction, whereas the Optivex patent claims are directed to statistical regression-based predictive methods for event detection; (ii) the data inputs and preprocessing steps described in the Optivex claims differ from those employed by the Verdana platform; and (iii) the Verdana system is designed for demand forecasting rather than anomalous event detection, which Breckenridge IP Group argues places the Verdana system outside the scope of the Optivex claims.
The Breckenridge IP Group analysis is currently in draft form and has not been finalized. While the preliminary conclusions are favorable, we note that the analysis has not been subjected to the level of rigor that would be applied in a formal invalidity or non-infringement opinion rendered in connection with litigation, and the draft status limits its utility as a definitive risk assessment tool.
Risk Assessment: Low-to-Moderate. Based on our review, we assess the risk associated with the Optivex claim as low-to-moderate. The claim appears to be an early-stage assertion that has not progressed to litigation and for which the claimant has not provided detailed technical support. The preliminary non-infringement analysis is directionally favorable. However, we note the following risk factors: (i) the Optivex patent covers a broadly-worded method claim in the supply chain analytics space, and broad patent claims can be difficult to design around if interpreted expansively; (ii) the patent assertion landscape in the AI and supply chain technology sectors is active and evolving; and (iii) even a meritless patent infringement suit can impose significant costs and distraction, particularly on a growth-stage company.
Recommendations:
(i) Pinnacle should request that Verdana obtain a finalized non-infringement opinion from Breckenridge IP Group LLP (or equivalent qualified patent counsel) as a condition to closing or as a deliverable concurrent with closing. The finalized opinion should address all asserted claims of the Optivex patent and provide a reasoned basis for the non-infringement conclusion.
(ii) The Stock Purchase Agreement ("SPA") representations and warranties should include a specific IP non-infringement representation with appropriate disclosure of the Optivex claim in the disclosure schedules. The representation should cover both actual and, to the Company's knowledge, threatened IP infringement claims.
(iii) Pinnacle should consider whether an IP-specific indemnity provision in the SPA is warranted, providing the Investor with indemnification for losses arising from the Optivex claim or any related IP infringement claim that materializes. Alternatively, or in addition, Pinnacle may wish to negotiate an escrow or holdback mechanism to provide a source of funds for potential IP defense costs, although we note that this is more commonly seen in acquisition transactions than in minority growth equity investments, and the current risk level may not warrant such a mechanism.
(iv) The transaction documents should include a covenant requiring Verdana to promptly notify Pinnacle of any developments in the Optivex matter, including any further correspondence, claim chart submissions, or filing of litigation.
B. Other Disputes
Based on our review of the data room materials, management representations, and the Company's litigation docket (data room Section 8), we did not identify any other pending or threatened intellectual property infringement, misappropriation, or other IP-related claims against the Company. We also confirmed that the Company has not asserted any IP infringement claims against third parties and has no prior history of IP-related litigation.
Sandra Mikkelsen confirmed in both the June 22 diligence call and the written Q&A responses that, other than the Optivex matter, the Company has not received any cease-and-desist letters, demand letters, or other communications alleging IP infringement, misappropriation, or other IP-related claims. This confirmation is consistent with our independent review of the data room materials.
X. INBOUND LICENSE AGREEMENTS
Verdana licenses certain third-party software components and data services for integration with or use in connection with the SCX platform. We reviewed the material inbound license agreements identified in the data room and summarize our findings as follows:
(a) Cloud Infrastructure Services Agreement. Verdana's primary hosting infrastructure is provided by Adriatic Technologies ("Meridian Technologies") under a standard enterprise customer agreement. The agreement is governed by the Meridian Technologies Customer Agreement and applicable service terms, which are standard across Meridian Technologies's enterprise customer base. The agreement does not raise IP ownership concerns — Meridian Technologies's terms are clear that customers retain ownership of their content and applications hosted on the Meridian Technologies infrastructure. The agreement is terminable by either party upon notice, subject to standard notice periods, and does not contain change-of-control or assignment restrictions that would be implicated by the proposed Series D investment.
(b) Geospatial Data and Mapping Tools License. Verdana licenses geospatial data and mapping tools from a third-party commercial provider under a standard commercial SaaS license. The license is structured as an annual subscription, is non-exclusive, and permits Verdana's use of the licensed tools in connection with the SCX platform's logistics routing module. The license is renewable annually and does not contain change-of-control provisions that would be triggered by a minority investment. No IP ownership issues were identified.
(c) Natural Language Processing Library License. Verdana holds a perpetual, royalty-paid commercial license for a natural language processing ("NLP") library from a third-party AI software vendor. The NLP library is used in certain components of the SCX platform's analytics and reporting functionality. The license is non-exclusive, fully paid (one-time license fee with optional annual maintenance/support), and permits incorporation of the licensed library into Verdana's commercial SaaS product. The license does not contain assignment or change-of-control restrictions that would be implicated by the proposed transaction. No IP ownership issues were identified.