Disclaimer: This is a personal investment thesis written for my own records and published for educational purposes only. It is not investment advice. I invested $310.50 in Numurus via StartEngine in September 2025. All equity crowdfunding investments are highly speculative and carry a significant risk of total capital loss. Please conduct your own research before making any investment decision.
A note on timing: This memo is written from the perspective of September 2025, at the point of investment. A retrospective note at the end reflects on what has changed since.
Company Snapshot
Numurus is a Seattle-based technology company developing AI and automation software solutions for industrial and edge computing applications. Its flagship product is the NEPI platform (Numurus Edge Platform Interface) — an open-source, low-code operating system that enables engineers to build AI-powered robotic and smart sensing systems in weeks rather than years.
The company is raising on StartEngine under Regulation CF, targeting up to $1.235 million at a $10.78 million pre-money valuation.
Investment date: September 2025
Amount: $310.50 USD
Platform: StartEngine
Security type: Equity (Preferred Stock)
Shares received: 440
Price per share: $0.75
Company valuation at investment: $10,781,175
Minimum investment: $300
Total raised this round: $231,955 (closed December 2025)
The Wave: Why This Market, Why Now
In September 2025, the robotics and edge AI market was at an inflection point. The cost of robotic hardware had fallen dramatically over the previous decade — a Boston Dynamics-quality robot arm that would have cost $500,000 in 2015 was available for under $50,000 in 2025. The hardware problem was largely solved. The software problem was not.
Building AI applications for edge devices — robots, sensors, industrial equipment — remained extraordinarily complex. A small engineering team trying to deploy a smart sensing system needed expertise in device drivers, GPU programming, AI model management, sensor fusion, and user interface development. Most teams had one or two of those competencies. Very few had all of them.
NEPI was designed to solve this problem. By providing a plug-and-play operating environment with pre-built modules for the most common edge AI tasks, Numurus could reduce the development time for a new robotic application from years to weeks. The WESMAR case study — a marine sensor company that used NEPI to develop an AI-enabled sonar product in five months — was the clearest demonstration of the value proposition.
The Industrial AI and Edge AI markets were projected to reach $50+ billion by 2030. More importantly, the demand was already there: energy companies, manufacturers, and defence contractors were all actively looking for ways to deploy AI at the edge without building the infrastructure from scratch.
Why now? The combination of falling hardware costs, rising demand for edge AI applications, and the absence of a dominant open-source platform created a window where a well-positioned early mover could establish the standard. NEPI's open-source model was a deliberate strategy to build the developer community before the commercial model scaled — the same playbook that Red Hat, Elastic, and HashiCorp had used to build durable enterprise software businesses.
The Surfer: Team and Traction
Jason Seawall, CEO and founder, had previously sold BlueView Technologies to Teledyne — a direct exit in the marine sensing and underwater robotics space. This was not a first-time founder in an unfamiliar market. He had built and sold a company in the exact domain where NEPI was being deployed.
The traction at the time of the raise:
- $4 million in lifetime revenues — meaningful for a company at this stage and valuation
- Strategic partnerships with NVIDIA, Qualcomm, Vecow, and Sidus — four of the most important names in edge AI hardware
- $2 million in prior VC funding (2022) — institutional validation before the crowdfunding raise
- 76% gross margin — a strong indicator of software-native economics
- Open-source community — NEPI was available on GitHub, creating a developer adoption pathway that commercial licensing models cannot replicate
The NVIDIA and Qualcomm partnerships were the most significant signals. Both companies have rigorous partner qualification processes. Being listed as a validated software partner for their edge AI hardware means Numurus had cleared technical and commercial due diligence that most startups at this stage have not.
The Moat: Defensibility
Numurus' competitive position rested on three overlapping advantages.
Open-source flywheel. NEPI was available as an open-source platform, which meant developers could adopt it without a procurement process. As the developer community grew, the platform accumulated integrations, bug fixes, and use-case documentation that made it more valuable for commercial customers. This is the same dynamic that made Linux, Kubernetes, and TensorFlow dominant in their respective categories.
Hardware partner integrations. The NVIDIA, Qualcomm, and Vecow partnerships meant NEPI was pre-validated on the most widely deployed edge AI hardware platforms. A competitor would need to rebuild those integrations from scratch — a multi-year process.
Domain expertise. Jason Seawall's background in marine sensing and underwater robotics gave the company deep credibility in one of the most technically demanding edge AI applications. The WESMAR case study was not a marketing exercise — it was a demonstration of what the platform could do in a genuinely difficult environment.
The moat scored lower than it might because the open-source model is inherently double-edged: the same openness that drives adoption also means that a well-funded competitor could fork the codebase and build a competing commercial product. The defensibility is in the community, the partnerships, and the founder's domain expertise — not in proprietary code.
The Numbers: Valuation and Position Sizing
At $10.78 million, the valuation was modest. The company had generated $4 million in lifetime revenues, had $2 million in prior VC funding, and had strategic partnerships with four major hardware companies. A 2.7× lifetime revenue multiple for a company with 76% gross margins and a defensible open-source moat is not expensive.
The concern was the current revenue run rate. FY2024 revenue was $33,700 — a significant step down from the lifetime revenue figure, which implied that revenue had been front-loaded in earlier years. Monthly burn of $20,900 against $33,700 in annual revenue meant the company was burning more than it was earning, with only 2.4 months of runway at the time of the raise.
This was the primary risk factor. The company was raising to survive in the near term, not to accelerate growth from a position of strength. That is a different proposition from a company with 12+ months of runway raising to capture an opportunity.
I invested $310.50 — the minimum plus a small increment. The position was sized to reflect genuine conviction in the thesis (the open-source flywheel, the hardware partnerships, the founder's track record) while acknowledging the near-term financial risk.
Risk Factors (As I Saw Them in September 2025)
1. Runway risk. 2.4 months of runway at the time of the raise was the most significant near-term risk. If the StartEngine raise did not close successfully, the company would need bridge financing or would face existential pressure.
2. Revenue concentration. The lifetime revenue figure of $4 million suggested that a significant portion of revenue had come from a small number of large contracts. If those customers did not renew or expand, the revenue base could contract significantly.
3. Open-source competition. The open-source model drives adoption but also creates the risk of a well-funded competitor forking the codebase. ROS (Robot Operating System) was already a dominant open-source platform in robotics — Numurus needed to differentiate clearly from ROS's ecosystem.
4. Hardware partner dependency. The NVIDIA and Qualcomm partnerships were the core of the distribution thesis. If either company changed their partner programme or developed a competing software layer, the Numurus value proposition would be weakened.
5. Team size. At the time of the raise, Numurus was a small team. Scaling the commercial business while maintaining the open-source community would require hiring, which the raise proceeds were intended to fund.
Go/No-Go Decision (September 2025)
GO — $310.50 committed via StartEngine.
The founder's track record (prior exit in the same domain), the hardware partner validation (NVIDIA, Qualcomm), the open-source flywheel, and the modest valuation were enough to clear the threshold. The runway risk was noted and the position was sized conservatively to reflect it.
| Metric | At Investment (Sep 2025) | Current (Mar 2026) |
|---|---|---|
| Investment Amount | $310.50 | — |
| Shares | 440 | 440 |
| Price Per Share | $0.75 | ~$0.75 |
| Company Valuation | $10,781,175 | $10,781,175 |
| Estimated Value | $310.50 | ~$310.50 |
| Return | — | 1.0× (nominal) |
| Status | Active | Active |
All financial data sourced from the Numurus StartEngine campaign page, the KingsCrowd analyst report, the SEC Form C filing, and the P2P Market Data campaign summary. This article will be updated as new information becomes available.
