Why pay for an assessment instead of going straight to build?

Because the build will fail if the workflow is wrong. We've watched several teams skip this step and burn $1M+ on the wrong agent. Discovery is the cheapest insurance you can buy.

What if you tell us not to build?

Then we don't build. We've recommended 'don't' on roughly one engagement in four — usually because the workflow is upstream of where AI helps. We'd rather be honest than booked.

Who from our side needs to participate?

Two line-of-business owners, one data/IT lead, and someone with budget authority. About 6–8 hours of total time across two weeks.

Do we need to share data with you?

We work with synthetic and de-identified samples by default. If a deeper read is required, we'll scope an NDA-protected data slice with your security team.

Discovery Assessment

A paid, fixed-fee assessment that scopes the problem and sizes the payback before anyone commits to a build.

Scope this engagement Read how it works

SHAPE · Fixed fee · 1–2 weeks · 2 senior consultants

Analysts examining a client workflow through a giant magnifying glass, sizing payback against effort and risk before committing to a build.

THE PROBLEM

Most AI initiatives stall because the wrong thing was scoped first — usually the model, not the workflow. Discovery puts a payback model on the table before a line of code is written.

01HOW IT WORKS

The shape of the engagement.

01Map

STEP 1

Workflow archaeology

Two weeks shadowing the line of business. We instrument the process — handle times, error rates, escalation paths, and real cost-per-transaction.

02Score

STEP 2

Payback model

Every candidate use-case scored on value, effort, risk, and operational readiness. We hand back a build-vs-buy-vs-don't recommendation per workflow — not a generic AI strategy deck.

03Memo

STEP 3

Decision memo

Reference architecture, payback model, and an executive memo your committee can act on. If the answer is 'don't build,' we say so — and stop the meter.

WHAT WALKS OUT THE DOOR

Decision memoReference architecturePayback modelUse-case backlog (scored)

02ACROSS INDUSTRIES

How discovery assessment applies in your industry.

SELECT SECTOR ↓

HOW IT APPLIES

Assess agentic workflows in middle-office ops, KYC refresh, dispute processing, and credit memo drafting. Carry over the model-risk lens early so it doesn't bottleneck production.

WHAT TRIPS PEOPLE UP

Model risk approval can take longer than the build. We scope the MRM track in parallel from week one.

EXPLORE INDUSTRY

ILLUSTRATIVE ENGAGEMENT

PAYBACK SIZED

$14.2M / yr

Tier-1 retail bank · agentic dispute triage

03FAQ

What teams ask before signing.

04OFTEN PAIRED WITH

What usually comes next.

03BUILD

Agentic Automation

Production-grade agents and workflows on your stack — harness, orchestration, tools, MCP, and evals.

02SECURITY

AI Security & Governance

Threat modelling, data-loss prevention, model risk, and audit trails for LLM and agent-based systems.

05IMPLEMENTATION

Implementation Services

End-to-end rollout — integrations, data plumbing, observability, validation harness, and rollback.