Straight answers, on the record.
The questions that come up on every call, written out in one place. Nothing here is off-limits, and nothing here is pitched. If a question is missing, send it to hello@airframe.ai; we answer directly, and we add it here if it helps the next reader.
About Airframe
What we are, who we are, and where we sit relative to the analyst firms.
What is Airframe, in one paragraph?
Airframe is the system of record for enterprise AI. We run a live registry of 17,000+ tools and 114,000+ deployment case studies, then write a three-format brief (Stack Audit, Vendor Deep Dive, Renewal Briefing) on what peer companies are actually running, replacing, and paying. The brief is free. The work carries forward across five surfaces: decisions enter Research, become procurement events in Renewals, become deployments tracked in Registry and rolled out team by team in Transformation, and stay live in Context, the ambient workspace inside Claude, ChatGPT, Cursor, and Windsurf. Together they move AI transformation from guesswork to a registered, measurable operating layer.
What does "system of record for enterprise AI" actually mean?
Most companies already have a system of record for people, for money, and for customers, and nobody runs a serious organization without them. AI does not yet have one, which is why every board deck this year is asking the same three questions and getting different answers from each team. Airframe is the layer that holds the answers in one place: every AI vendor under contract, every deployment in flight, every owner accountable for an outcome, every renewal on the calendar, every brief written about a category your team is buying into. The registry sits underneath the work the way a general ledger sits underneath finance, and it is built to outlast any single tool, model, or vendor cycle.
How is Airframe different from the analyst firms?
The incumbents publish quadrants and waves quarterly, built largely on vendor briefings and practitioner surveys. Airframe runs a live registry grounded in peer behavior rather than vendor self-attestation, so the picture your team sees is what is actually deploying right now, not what vendors are pitching. We do not run paid briefing programs, we do not sell category-inclusion placement, and we operate with zero vendor conflicts. The brief your team receives is built to one purpose: make your renewal meeting start from the truth.
Is Airframe a tool, a service, or a team of people?
All three, and the lines are drawn deliberately. Research, Registry, Renewals, and Context are software, available on a subscription. Transformation is software that ships with a light deployment service, because rolling agents into production needs an audit layer and a plan from day one. Senior operators run alongside your central team during the quarters when you are standing up the AI function or scaling it outward across marketing, finance, operations, engineering, and legal. Your internal team keeps the reins in every model. Airframe is the layer underneath.
What about token spend and agent spend; is that what Airframe tracks?
Token spend and agent spend are the questions the agent era is going to keep asking, and we hear them in every CFO conversation and every board prep cycle. Today the registry covers AI vendor inventory, deployment ownership, and deployment outcomes; it is the system of record where those answers will live as the underlying telemetry catches up. Most finance teams arrive at the same place, which is that they cannot answer "what are we spending on AI and what is it producing" until there is a registered, owned inventory of what is actually deployed. That inventory is the work we do today, and it is the foundation any honest answer to the spend question will eventually sit on.
Who is behind Airframe?
Airframe is built by a group of operators and ex-founders. Across prior companies, the team has raised hundreds of millions of dollars and has spent the last decade buying, selling, and deploying the software categories the registry now tracks. The work is privately underwritten, the company is structured to operate without the time pressure of a venture-funded growth path, and we do not take vendor sponsorship, syndicated research budget, or paid placement of any form. If you want the long version, we are glad to walk you through it on a call at hello@airframe.ai.
The engagement
What it costs, how long it takes, who owns the work, and how it shapes against where you are.
Is a brief really free, and what does everything else cost?
Briefs are free. Every Stack Audit, Vendor Deep Dive, and Renewal Briefing is free to request, and the brief is how you see what we see before anything else. Most organizations use it without going further. The working subscription bundles Research (direct registry access, named analyst, private briefings), Registry (your private vendor and ownership inventory), and Renewals (the contract calendar and fair-range pricing), priced by cohort tier on an annual basis. Transformation and Context (the rollout plan, the agent audit layer, and the ambient workspace inside Claude, ChatGPT, Cursor, and Windsurf) sit on top of that subscription, with a usage component tied to agents in production. Senior-operator engagements are optional and scoped per engagement. We do not take vendor sponsorship, affiliate fees, or paid placement.
How long does an engagement take?
A brief takes 72 hours from request in the three formats above. The first Stack Audit of a new cohort member usually lands inside ten business days, because we want to spend the first week reading your integrations and your peer cohort before we write anything you will hand to a board. Registry, Renewals, and the Context workspace go live the same week as signing. Transformation rollouts are scoped to the surface area we are protecting and typically take two to four weeks from intent to first agent in production with the audit layer live. Senior-operator engagements are scoped per engagement and usually run in twelve-week windows against a named deliverable.
What does Mode 01, Mode 02, and Mode 03 mean?
The three modes are how an engagement is shaped against where your AI function actually is. Mode 01 (Stand Up) is for organizations without a central AI function yet, where the first six weeks are spent building one: registry live, renewal calendar in place, the first two or three deployments owned by named people. Mode 02 (Scale) is for teams that already have a function and need to push deployments outward across marketing, finance, operations, engineering, and legal, often against EBITDA-tied targets. Mode 03 (Run) is the always-on cadence afterward, with the registry, renewal calendar, and Context workspace maintained as a live operating layer rather than a one-time project. The mode is named in the engagement letter, and it is the same operating system in each case, scoped differently.
Who owns the work, your team or ours?
Your team. Every engagement is built so the registry, the renewal calendar, the deployment plan, and the agent audit layer live inside your organization and stay there if we walk away tomorrow. Senior operators from Airframe sit alongside your central AI team during the quarters when the lift is heaviest, but the named owner of every deployment is on your side of the table, and the registry is yours. We are explicit about this because the alternative, which is consulting work that leaves the binder on a shelf, is exactly the failure pattern this category has been stuck in for a decade.
How does this work for a private-equity firm versus a single enterprise?
Same operating system, different unit of analysis. For a single enterprise, the engagement runs against one registry, one renewal calendar, and one set of teams. For a private-equity firm, the engagement runs across the value-creation portfolio: portcos opt in one at a time, each with its own registry inside the firm-level view, and the operating partner sees the cross-portfolio picture (what is being deployed, what is being replaced, where the EBITDA-tied targets are tracking) without losing the per-portco detail. The three modes (Stand Up, Scale, Run) map cleanly onto a portco's stage, which is why a single firm can run all three at once across different holdings.
Who is the founding cohort for?
The founding cohort is a small group of Fortune 1000 finance and technology leaders who want to shape the platform as it comes online, in exchange for pre-launch terms that carry forward. Membership is by review. We are looking for organizations with a meaningful AI renewal cycle ahead of them in the next four quarters, and a leader willing to work in the open with us. If that sounds like your next twelve months, apply here.
Methodology and data
Where the registry comes from, how fresh it is, and how your data is held.
Where do the 17,000+ tools and 114,000+ case studies come from?
The registry is built bottom-up from primary sources, not from vendor self-attestation. Each tool entry begins with the vendor's own product surface, then layers in deployment evidence drawn from public case studies, regulatory filings, hiring patterns, integration footprints, and the renewal-cycle behavior we see across the cohort. The 114,000+ deployment case studies are individual, named instances of a tool being put into production at a named organization, with the buyer team, the use case, and (where disclosed) the outcome attached. The methodology is laid out in full at registry methodology, including how categories are defined, how duplicates are reconciled, and where we draw the line on what counts as a deployment versus a pilot.
How fresh is the data?
The registry is live, not a quarterly snapshot. Vendor entries are touched whenever new evidence lands (a funding round, an executive move, a deployment, a category-relevant filing), and the public-facing parts of the Index are rebuilt on a continuous cadence rather than a fixed publication date. For cohort members, the Renewals calendar inside your private workspace is updated as contracts move and as peer-cohort pricing data refreshes, so the fair-range you see on a renewal meeting is what we see today, not what we saw last quarter. The 1974–2026 long-form Register, which underwrites the wave-collapse thesis, is re-verified against the source data file on every editorial pass.
Do we need to share our contracts, and is our data safe?
No on the first, yes on the second. A vendor list and a few renewal dates are enough to begin, and the fair-range pricing we surface is drawn from peer cohort patterns rather than from your paper. If you do choose to share contracts, everything you upload sits inside a private Context workspace scoped to your organization, and nothing flows back into the public registry or another customer's view, ever. SOC 2 Type II is in progress and will be signed before the first Transformation rollout puts agents into production. If your stack touches regulated data, we run the ingest through a secure data room with access logged and permissioned by role, and we will sign the paper your security team requires before the first byte moves.