Service - Discovery phase

TRACE Discovery: a written read on your AI readiness

TRACE Discovery is a 1 to 2 week assessment for enterprises in regulated industries who want to know, on paper, where AI is genuinely ready to ship in their operation. We map readiness, regulatory exposure, and the use cases with the highest leverage, then hand over a prioritised roadmap and a fixed scope for the Build phase. No code is written.

1-2 wk

End-to-end duration

3-6

Use cases scored against TRACE

Written readiness report, board-ready

Pricing surprises in the follow-on quote

Book a discovery call Read the TRACE methodology

Section 01

01.What does TRACE Discovery deliver?

Discovery delivers a single written report and a prioritised roadmap. The report scores 3 to 6 candidate use cases against the four TRACE pillars: Trust (regulatory and data-residency exposure), Readiness (data quality, workflow maturity, integration surface), Architecture (the abstract system needed to ship the use case), and Citations and Evidence (what an auditor or regulator would expect on day one). The roadmap orders the use cases by leverage versus risk and proposes a Build sequence with named owners on your side. The fixed scope for the Build phase is the artifact that lets a CFO or CIO commit budget without further rounds of estimation.

Section 02

02.Who is TRACE Discovery for?

It is for CIOs, CTOs, and Chief Risk Officers in regulated industries (legal, debt collection, insurance, banking, healthcare, logistics) who have a board mandate to deploy AI but no ranked, defendable plan. It is for organisations that have already tried, and shelved, at least one AI pilot, and want to understand why before they spend again. It is for in-house teams who want a second pair of senior eyes on a roadmap that internal politics will not let them write honestly.

Section 03

03.What is not included in Discovery?

Discovery does not ship code. It does not run prompts in production. It does not promise vendor recommendations: the architecture descriptions are vendor-agnostic by design, because the right vendor is selected during the Build phase against your numbers. It does not produce a slide deck as the headline deliverable. The headline deliverable is a written report and a fixed-scope statement of work, in that order.

Section 04

04.How does it compare to a typical AI strategy engagement?

A typical strategy engagement runs 6 to 12 weeks, produces a slide deck and a heat-map, and does not commit to a build scope or a number. TRACE Discovery runs 1 to 2 weeks, produces a written report and a fixed scope, and the consultancy that wrote it is the same one that will deliver the build. The accountability runs through the same team end to end. We also charge a flat fee against a defined deliverable list, not against time and materials.

Section 05

05.What turns Discovery into a Build engagement?

The Discovery report ends with a recommended Build sequence. The first item in that sequence becomes either an AI Sprint (4 to 6 week proof on a controlled cohort) or an AI Lighthouse Build (8 to 12 week production deployment), depending on what the readiness score allows. We quote against the fixed scope inside the report. You retain the right to take the report to another implementer.

Methodology mapping

06.How this SKU sits inside the TRACE methodology

This engagement is the Discovery phase of the Impetora delivery model. Discovery is the first of three phases (Discovery, Build, Operate). Its job is to make the next two phases boring.

Trust

We classify the candidate use cases against EU AI Act risk tiers and map data flows for GDPR alignment, before any architecture choice is made.

Readiness

Real-volume sampling of source data, baseline measurement of current process outcomes, and an honest read on integration surface.

Architecture

An abstract reference architecture per use case: ingestion, retrieval, model layer, evaluation, observability, with no vendor lock-in baked in.

Citations and evidence

The deliverable is a written report. Every claim links to the document, log, or interview that supports it. A board can defend it.

Engagement model

07.What happens, week by week

01Day 1
Kick-off and document pull
Two-hour kick-off with the executive sponsor, then a request list of policies, prior pilot post-mortems, regulator correspondence, and the workflows on the table.
02Days 2-6
Workshops and sampling
Four to six 90-minute workshops with operations, IT, risk, and legal. Real-volume sampling of the candidate workflows.
03Days 7-10
Report drafting and walk-through
We draft, send for review, then walk the report through the steering group. The fixed-scope statement of work for the Build phase ships alongside.

Scope of work

08.Inputs we need from you. Outputs we ship.

Inputs we need

From your team

Executive sponsor with the calendar to sign off on scope at the end of week two
Read access to representative samples of the candidate workflows
Existing policies on data residency, AI use, and procurement
Any prior AI pilot post-mortems, even if they did not ship
30 to 60 minutes each from operations, IT, risk, and legal

Outputs we ship

Concrete deliverables

Written readiness report (typically 25 to 40 pages) with the four-pillar scoring per use case
Prioritised Build roadmap with sequencing, named owners, and an honest risk register
Fixed-scope statement of work for the recommended first Build engagement
Workshop summaries with explicit decisions and open questions, dated and attributed
Vendor-neutral reference architecture sketches per use case

Honest scoping

09.Who this is not for

We turn engagements down when the fit is wrong. If any of these match, a different SKU, or a different partner, will serve you better.

See the full list of fit signals we screen against

Organisations that have already chosen a vendor and need an implementer, not a diagnostic
Teams without an executive sponsor who can commit a budget envelope at the end of two weeks
Use cases where the workflow itself is undefined; we can score AI readiness, not invent your process
Buyers looking for a slide-deck strategy artefact rather than a working scope of work

Frequently asked questions

How long does Discovery take, end to end?

One to two weeks of elapsed calendar time. We schedule the workshops in the first week and use the second week for drafting, review, and the final walk-through. Faster than two weeks is technically possible on a single use case but risks the report being thinner than a board needs.

Do you sign an NDA before kick-off?

Yes. We sign your NDA, or we provide ours. Mutual non-disclosure is in place before any document changes hands. We also sign a Data Processing Agreement if any personal data crosses the engagement perimeter.

What if the report says we are not ready?

It sometimes does. The honest answer is named owners on remediation, a re-Discovery date 90 to 180 days out, and no Build engagement until the readiness gates are met. We would rather walk away from a Build that will not ship than collect a fee against a project that will not.

Can we take the report to a different implementer?

Yes. The report is yours. The fixed scope inside it is written so any competent implementer can quote against it. We do not gate the deliverable on signing a Build contract with us.

Is the Discovery deliverable available in languages other than English?

Yes. We deliver in English by default and in Lithuanian, German, French, or Spanish on request. Workshops are run in the language the team prefers; the written report is delivered in the agreed language.

Do you bill by the hour or by the engagement?

By the engagement. The Discovery fee is a fixed amount tied to the deliverable list, agreed in writing before kick-off. There is no time-and-materials variant of this SKU.