From Intent to Yield: Why We Partnered with Jama

A View from the Watchtower · Part Three

Requirements aren't a document you sign and file. Inside EDA 3.0 they are the living spine of the program — and we're proud to welcome Jama Connect to the portfolio.

We have spent two posts on requirements because we believe something the industry has been slow to say out loud: in EDA 3.0, requirements are not the paperwork around the design. They are the design's first layer — and the thread that holds every layer after it together.

The six layers, and the one everyone skips

Our thesis is simple. EDA 2.0 was tool-centric, siloed, and human-driven — value measured in how fast a single tool ran. EDA 3.0 is AI-native lifecycle orchestration from intent to yield, across six layers: L1 Intent and Requirements, L2 Architecture and Modeling, L3 RTL and Firmware, L4 Verification, L5 Yield and Physical, and L6 AI Orchestration. Value is measured in outcomes — time-to-market, yield, margin — not tool speed.

Almost all of the industry's attention, and nearly all of its tooling, lives in layers two through five. Layer one — what the chip is actually supposed to do, for whom, and how we will prove it — is the layer everyone assumes someone else owns. That is exactly why programs drift. A simulator can tell you the design does what the RTL says. It cannot tell you the RTL says what the stakeholder needed. Only the requirements layer can — and only if it is alive.

One unbroken line, intent to yield

Picture a single thread running up the stack. It starts as a stakeholder need at L1, becomes an architectural decision at L2, an implementation in RTL or firmware at L3, a verification plan and result at L4, a physical and yield outcome at L5, with AI orchestrating the whole flow at L6. When that thread is continuous, a change at any layer is instantly legible everywhere else. When it is broken — when L1 is a PDF nobody has reopened since kickoff — every layer beneath it is flying on stale intent.

Live Traceability is that thread. It is why Jama Connect is the right anchor for Layer 1: it does not merely store requirements, it keeps them connected to everything downstream as the design moves. That is the difference between a requirements document and a requirements spine.

Why AI Tech Sales partnered with Jama

We are an EDA 3.0 lifecycle-orchestration firm. Our job is to connect the layers, not to sell another point tool. We had a strong Layer 1 argument and a gap in the portfolio: the category-defining platform for living requirements. Jama Connect is that platform — and we are proud, finally, to welcome it to the AI Tech Sales portfolio.

The fit is almost uncomfortably good. Jama is Portland-built, like us. It owns and named the Live Traceability category. It has just shipped a semiconductor-specific solution. And it is leaning hard into exactly the AI-native direction our thesis predicts — making requirements the trustworthy context layer beneath AI-assisted engineering. When the platform you would have drawn on a whiteboard turns out to already exist, you stop drawing and start partnering.

What it means for our customers

For the teams we work with, the translation is concrete. Requirements become the living spine of the program — queryable, connected, and current, instead of a document you sign and file. The plan keeps pace with the design — a change anywhere surfaces everywhere, before it becomes a respin. Safety, security, and quality become one system — properties of a single traceable program, not three separate audits. And AI finally has trustworthy intent — the one thing AI-assisted engineering cannot manufacture for itself: a traceable source of truth to ground it.

Book a briefing →

The Jama Series · A View from the Watchtower

Part 1 — The Spreadsheet Was Never the Plan
Part 2 — Functional Safety Was the Front Door, Not the House
Part 3 — From Intent to Yield: Why We Partnered with Jama (you are here)