The AI StudioSample engagement · what we build
Sample · Client details anonymized
The AI StudioSample engagement · what we build
Sample · Client details anonymized
We evaluated a live-entertainment group end to end - fifteen people, six workflows, every system - found 57 concrete opportunities, and built a private AI platform to go and get them. This is the evaluation, the proposal, both statements of work, and the financial impact estimate we delivered, with the client’s name, people, and confidential figures removed. Thirty days of discovery, then ninety days of build: nothing to all 57 in 120 days.
Every AI Studio engagement runs the same way: we learn how the business actually works before we recommend anything, we propose only what the evidence supports, and we write down the test that says each piece is done.
And the far end is not preordained either. The evaluation decides the proposal, not the other way round: for this client the evidence pointed to a private AI platform - the AI Studio. For a different business it can point somewhere else entirely, and that is what we propose instead.
Thirty days of discovery: who we interviewed, the systems & data map, six workflow maps drawn end to end, findings by area, what we measured, the full 57-item inventory with confidence ratings, the roadmap, and the governance model the client’s own team set.
Read the EvaluationA private AI platform, owned outright, on the client’s own server - with the first eleven builds chosen straight from the evaluation. Two tiers: the Platform at $300K (platform + 11 builds covering 14 of 57) and the Full Studio at $600K (all 57, done for you). Every one of the 57 assigned a lane. And the Studio itself is what this client’s evidence called for - your evaluation may call for something different, and that is what we would propose.
Read the ProposalThe build specification for the Platform tier: architecture, connectors, security, two milestones with acceptance criteria and payments attached, the data-accuracy verification loop, and a written acceptance condition for each of the eleven builds.
Read Statement of Work AThe second statement of work, written after the first build was already running on their data - which is why it can be this specific. It finishes the other 43 of the 57 and turns the Studio from an analyst into an operating platform: department agents, plain-English automation, and approved actions inside their own systems with a receipt for every one. Thirty days, 37 rows, each with its own acceptance condition - and a section naming the four things we would not promise.
Read Statement of Work BWhat the 57 are worth, in dollars. The first five builds estimated individually with the arithmetic shown, the remaining fifty-two in aggregate, the collective outlook expressed against net margin, and an illustrative five-year model with the payback point marked. Conservative by construction: roughly twenty opportunities are carried at zero.
Read the Financial ImpactThese are working documents from a real engagement, not a brochure. The evaluation is written to the client’s leadership; the proposal and statement of work are the ones that followed from it. The method, the structure, the diagrams, the numbers your team stated, and the way scope and acceptance are written are all as delivered.
What we removed: the client’s name and its operating units (they appear as the Touring division, the Concerts division, and the ticketing subsidiary); every person’s name (they appear by role); and confidential financial actuals. What we left in: the two build-tier prices, because a schedule without a number attached does not tell you much. Read them as the shape of a deal this size rather than a rate card - every engagement is priced from what its own evaluation finds. Vendor and system names stay, because they are what make a connector plan concrete.
A live-entertainment group producing more than 1,000 shows a year across a touring division, a concert-promotion division, and a separate ticketing company; roughly 160 people; finance on a major ERP and spend platform, ticketing across two vendors and a daily-count platform, a 60-million-row ticketing warehouse, and an operations database for offers.
This is not a concept bundle. The engagement these five documents describe is being delivered today. The evaluation is complete. The first statement of work allowed 133 days for the platform and its eleven builds; it was in front of the leadership team, running on their own systems and their own data, 43 days after the go-ahead. The second statement of work, which closes out the remaining 43 opportunities, is written and costed.