Diagnose
Get the facts before you build anything. Diagnose programs run for one to three weeks and end with a report, a score, or a roadmap. No code is written.
What this tier is for
These programs are for teams that want to understand their situation before committing to a build. Each one is fixed scope, runs for one to three weeks, and ends with something concrete, whether that is a reliability score, a roadmap, a cost breakdown, or an audit. You leave with facts. No code is written in any Diagnose program.
Diagnose programs
Scope
Maps your business and operations over 2–3 weeks to find where AI creates real value, models the ROI for each opportunity, and produces a prioritised roadmap.
What you get
Costed, prioritised AI roadmap with ROI model, board-ready
Scope
Reviews a live AI system over 2 weeks using real outputs, logs and traces, scored across reliability, evaluation coverage, cost and observability.
What you get
Scored reliability report with a prioritised remediation plan
Scope
Analyses your existing workflows over one week to score each for automation candidacy and model the financial return for the top three.
What you get
Workflow-to-agent map, automation candidacy scores and top-three ROI models
Scope
Reviews your engineering team’s development process phase by phase over 2 weeks, with a DORA baseline and an assessment of AI tooling across each phase.
What you get
Phase-by-phase SDLC audit, tooling recommendation, adoption roadmap and DORA baseline
Scope
Analyses your token and GPU spend across models and endpoints over 2 weeks, identifying where costs are highest and what optimisation options are available.
What you get
Token and GPU spend analysis with a costed optimisation plan
Which to choose
AI Opportunity & Roadmap Sprint
if you want a full business-level AI roadmap with ROI models and a sequenced plan your leadership team can act on.
if you have a live AI system and want to know how reliable it is, where the gaps are, and what to fix first.
if you want to know which of your workflows are the strongest candidates for agent automation and what the return looks like.
if you want to understand how AI-native your engineering process is and where your team should adopt AI tooling next.
if your AI inference spend is high and you want to understand exactly where it is going and how to reduce it.