We're early. The thesis is clear.
Two ways to help us build.
OpsATC.AI is at the build phase, bootstrapped and funded by paid design-partner pilots. We are not raising today. To get from here to a deployed, paying, named-customer product as fast as possible, we're recruiting on two fronts: design-partner pilots and a technical co-founder. If either is you, this page is for you.
For pilot prospects reading this page: the incoming technical co-founder commits to the architectural doctrines published on the Trust Center and Approach pages (read-only against systems of record, tenant isolation, and a commitment to amplify rather than replace your workforce) as a precondition of partnership. The next engineer won't change the doctrine. The doctrine is what the platform is.
From pilot LOI to a live deployment.
"Speed to launch" needs a shape behind it. Below: the path from a pilot LOI to The Captain live in a partner's stack, and the two fronts (pilot customers and a technical co-founder) we're working in parallel to get there.
Day 0 is the LOI with the first pilot partner, and the technical-co-founder search runs in parallel so the build can start cleanly.
A note for pilot partners reading this page. The technical co-founder commits to the architectural doctrines published in the Trust Center and the Approach as a precondition of partnership, including the read-only doctrine, tenant isolation, and our commitment to amplify rather than replace your workforce. The doctrines are settled; the team is being shaped around them, not the other way around.
Operations is the largest under-served software market in the AI era.
Product companies (OEMs), hub providers, contract manufacturers, distributors, logistics carriers, and high-mix specialty operations move trillions of dollars of physical goods every year on top of an ERP / S&OP / WMS / CRM / PLM stack that doesn't talk to itself. Foundation models that can reason, Model Context Protocol that lets AI read from systems of record auditably, and the observability tooling around long-running agent analyses all matured at the same time in 2024-2026.
The connective layer above the systems, the layer that's been a human clicking between forty browser tabs for fifteen years, is finally buildable. OpsATC.AI is built to be that layer, with a vertical specialization that horizontal AI players can't credibly deliver and an architectural commitment (read-only by default, cited responses, tenant isolation, no foundation-model training on customer data) that operations buyers actually require.
The founder's full thesis, the multi-industry pattern recognition behind it, and the connective-layer framing live on the Our Team page. The product, the architecture, and the comparison vs. alternatives live on the Platform page.
For pilot partners.
OpsATC.AI is openly searching for a small founding cohort of pilot partners, product companies (OEMs), hub providers, contract manufacturers, distributors, and the logistics carriers that move their product, who want a hand in shaping the product, an early seat at the table on roadmap, and the price hold and protections that come with being first. The faster we can put the platform inside a real operation, the faster the finished product gets to market. That makes the search public.
Who we're looking for
Product companies (OEMs), hub providers, contract manufacturers, distributors, logistics carriers, and high-mix specialty operators moving real physical product across multiple systems of record, typically several sites and a recognizable stack: ERP, S&OP / planning suite, WMS, TMS, CRM, PLM, EDI. Complexity matters more than size: a fast-growing hardware OEM running five connected systems with hard allocation calls is a better fit than a larger operation on a single ERP with simple flows. Operations-led, with a leadership team comfortable being early on a category-defining product. Industry-agnostic: hardware and storage OEMs, manufacturing, distribution, integration, logistics, hub services, medical devices, industrial automation, specialty.
What pilot partners get
A pilot fee, sized to the scope, scale, and integration footprint of your engagement, followed by a 1-year or 2-year commercial contract at a materially-discounted monthly rate locked for the full term. The pilot fee is credited in full against your first-year subscription fees when you convert, applied ratably across the year. It covers the pilot engagement itself rather than sitting as a deposit, so it isn't returned separately. If you decide not to continue, you keep the findings, the configuration, and a complete data export, and owe nothing further. On top: documented roadmap influence; direct founder relationship; change-of-control protections, source-code escrow, perpetual license to the version you're running; tenant isolation; no foundation-model training on your data. Full pricing posture and "what design-partner pricing buys you" detail on approach.html#pricing.
What we ask in return
Access to a real operating scenario inside your business. A small, focused stakeholder group across operations, IT, and a sponsoring executive. A focused pilot commitment to first measurable outcome, in weeks not months: cycle time, OTIF, exception MTTR, onboarding velocity, decision compression. Honest feedback during the pilot, in real time, so the platform is shaped by operators who actually run an operation. Outcome timing is projected and depends on your circumstances and the level of access granted.
Why this compresses time-to-launch
The fastest path from today to a deployed, named-customer product is to put real data, a real operation, and a real worst-week scenario in front of the platform from the first sprint. Pilot partners shorten that path by making the build target concrete. They tell us which connectors, which workflows, and which recommendation surfaces matter first. In exchange, they shape the version 1.0 they end up running.
Six questions a pilot partner should ask, answered straight.
How do you pick which workflow to start with?
We don't. You do. The first sprint is a scoping conversation where we name three candidate workflows in your operation, rate each against architectural fit and impact, and pick one together. The pattern: pick the workflow that hurts most weekly but doesn't yet have a competing in-flight tool. That's where The Captain shows value without political friction.
What if our ERP isn't in the five reference adapters?
695 platform adapters are scaffolded today; 141 are built to our Tier-1 rubric, written against each vendor's documented API contract with a typed client, a health-check ping(), and replay fixtures (including SAP S/4HANA, SAP ECC, Oracle Fusion ERP, Manhattan Active WMS, and Microsoft Dynamics 365 Business Central). Live-sandbox validation is per-adapter onboarding work and is in progress. SAP S/4HANA, New Relic, Confluence, and PagerDuty are proven live end-to-end today. Plus 8 generic adapters (JDBC, OData, OpenAPI, GraphQL, SFTP, SOAP, Webhook, X12) bridging anything not yet in the catalog. Activating one of the scaffolded targets is typically under a day of senior-implementer time per adapter (the codegen does most of it), and we confirm that estimate against your specific system during onboarding rather than promising it up front. If your stack is in the scaffolded set, we activate it during pilot kickoff. If it's truly novel (an on-prem fork, a custom DB schema), we scope a custom adapter into the pilot fee in writing before you sign.
Can we pause the pilot?
Yes. The pilot agreement includes a pause provision: written notice, immediate effect, no cancellation penalty. The pilot clock pauses; resumed days count from the same point. We have not run a pilot yet, so we have no pause history to characterise for you. The clause exists because an unrelated systems migration or a sponsor's absence should not cost you pilot days, and we would rather write that down now than negotiate it later.
What happens to our data after the pilot ends?
If you don't sign a commercial contract, your tenant is exported, your read-only credentials are revoked, and your tenant is archived and slated for deletion (a verifiable hard-deletion SLA is on the roadmap). You receive a copy of the KPI definitions you configured, the prompt templates configured for your operation, and the conversation audit log. You keep the work; we just stop running it. Full data-residency and exit detail on trust.html → portability & exit.
Do we have to be a named public reference?
No. Reference willingness is optional and never contractual. Named design partners go public only with explicit written consent, never as a marketing announcement. The first named partner will be a milestone, not a press release. Design partners who decline to be named keep every other commercial benefit (locked rate, pilot-fee credit, roadmap influence, direct founder access).
Who owns the IP of pilot-period configuration?
You own your KPI definitions, your prompt templates, your scheduled cadences, and your role-mapping. OpsATC.AI owns the underlying platform code and the vertical configurations that ship as part of the platform. Where your configuration informs the shipped vertical config (the "roadmap influence" benefit), we name your operation as the source in the changelog (with your consent), and you receive the same commercial protections design partners get on every shipped version that lands in production.
If your operation is ready to be early, let's talk.
Email [email protected] with the subject line Pilot partner conversation. Include a short description of your operation, your top operational pain right now, and a name we can call you by. You'll hear back within one business day. The full set of inputs that make a tailored demo possible is on the demo prep page, but the first conversation only needs an introduction.
Bootstrapped, and funded by paid pilots.
OpsATC.AI is bootstrapped and funded by paid design-partner pilots: founder-funded to date, with pilot revenue as the next milestone. We are not raising outside capital today.
A Series A is possible later, if and when the Board directs one, and on the back of named, paying customers. But there is no active round, no projections, and nothing on this page is an offer to sell or a solicitation to buy securities.
For technical co-founders & founding engineers.
OpsATC.AI is looking for the technical leader who will build the platform alongside the founder. Co-founder economics for the right person; senior founding-engineer terms for the right person who isn't ready for full co-founder. The technical scope is unusually rich for a pre-seed company because the architecture commitments are already settled.
The technical scope
Anthropic Claude foundation models as the agent reasoning layer. Model Context Protocol (client + server) as the integration fabric across SAP, Oracle, Kinaxis, Manhattan, Salesforce, Windchill, ServiceMax, Snowflake, Databricks, EDI/X12, and custom MCP endpoints. A multi-tenant SaaS architecture with strong per-customer data isolation. An operational knowledge graph kept in sync with systems of record. The Process Intelligence Engine that surfaces bottlenecks and quantifies impact. Five role-aware portals on top.
The kind of person we're looking for
Someone who has shipped a production AI / agent / orchestration system before, ideally one that touched real systems of record under real audit constraints. Someone fluent in Python, TypeScript, modern cloud infrastructure, and at least one ERP or supply chain platform. Someone who has worked on enterprise multi-tenant SaaS and understands the discipline that takes. Someone who reads MCP, agentic-tool-use, and integration spec docs for fun. Someone who values architectural commitments over feature velocity.
What you'd own
The technical architecture, end-to-end. The MCP connector fabric and the SDK that lets new connectors ship in days, not months. The agent layer (The Captain) and the safety / HITL machinery underneath it. The audit, observability, and trust posture. The hiring of the next two to four engineers as the team grows. The technical voice in design-partner conversations.
What you'd get
Founder-grade equity (for the co-founder path) or founding-engineer equity plus market salary (for the founding-engineer path). Direct partnership with an operator-founder who has run global supply chain operations and will be on every customer call with you. A category-defining problem in a market that's actually big. The ability to do your best work without a large company's politics, and the architectural quality bar to make the work worth doing.
If this is the work you've been waiting to do, let's talk.
Email [email protected] with the subject line Technical co-founder conversation. Include a short description of the most relevant thing you've built, your fluency with Anthropic Claude / MCP / agentic systems, and what you'd want to talk about in a first call. You'll hear back within one business day.
Why the conversation is worth having.
What's already in place is what compresses the timeline once the technical depth lands: operator-founder credibility, architectural commitments, brand, engagement methodology, and an active design-partner pipeline. If you look at what's been built, the conversation is worth having.