A THOUGHT EXPERIMENT
What would you do with this team?
Before we talk about your roadmap, hold this picture. This is roughly what a year of agent capability buys at today’s model pricing.
1,000
Summer students
At about $0.25 an hour. Each one remembers every conversation you have ever had with them.
100
PhD-level researchers
At about $25 an hour. Each one focused on any topic you assign, on demand.
10x
University graduates, per existing employee
Working alongside your team, on demand, at whatever hour the work needs doing.
How would you structure your departments? Your teams? Your workflow?
Most organizations have never designed for this. That is what the AI Readiness Assessment is for: it turns the thought experiment into a costed, sequenced plan for your business.
Illustrative, based on published model pricing at the time of writing.
You keep what we build. Designs, standards, code, documentation and the measurement baseline transfer to your team, and ownership sits with your technical leadership. Nothing we deliver needs us to keep it running; where you want us to stay, that is a published engagement with its own scope, not a dependency.
THE JOURNEY
Fifteen engagements across five stages.
An engagement has a published deliverable, a duration and a named delivery owner, and it has its own page. A capability is something we do inside those engagements; it is named here so you can see the whole surface, and it does not carry an availability marker.
Assess
Before anything is built, agree what is worth building.
QW
AI Readiness Assessment (AIORA)
Leadership immersion to a board-ready twelve-month roadmap with three to five costed pilots.
QW
Microsoft 365 Copilot Readiness Check
Permissions, data exposure, access and content control, checked before you turn Copilot on.
QP
Platform Foundation Review
A diagnostic on the estate you already have, ending in a hardening plan.
Ready
The unglamorous work that decides whether everything after it succeeds.
QP
Data Readiness
A governed data estate your AI can actually retrieve from, on Microsoft Fabric or Databricks.
QP
AI Platform Engineering
The harness your AI runs in: landing zone, guardrails, audit trails and Agentic FinOps.
QA
AI Solution Design
Architecture, model selection, integration approach, containment plan and acceptance criteria, agreed before code is written.
QP
Data architecture and modernization
Scattered sources consolidated into a target-state estate. Delivered inside Data Readiness.
QP
AI landing zone and guardrails
Reference architecture, infrastructure as code, CI/CD, role-based access, approvals, audit and rollback. Delivered inside AI Platform Engineering.
QW
AI policy and responsible use
Acceptable use, data rules and approval paths, in language a board will accept. Delivered inside the AI Readiness Assessment (AIORA) and AI Deployment.
Deploy
Where it stops being potential.
QW
AI Deployment
From licenses to measured adoption in four weeks. Editions for Microsoft 365 Copilot, Claude, ChatGPT, Gemini and Cursor.
QA
Process Re-engineering
One painful process rebuilt around agents, with impact measured against a baseline.
QA
AI SDLC
A consistency layer on your existing stack, so the team ships faster and not just individuals.
QA
Agent and Application Build
A working agent or custom application, built, governed and deployed.
QA
Pilot to production
A stalled pilot taken to a governed, supportable deployment. Delivered inside Agent and Application Build.
Adopt
A deployed tool nobody uses has cost you money.
QW
Adoption and change enablement
Moving sanctioned AI into daily workflows. Delivered inside AI Deployment and the AI Adoption Managed Program.
QW
Prompt libraries and clinics
Facilitated sessions and curated libraries that keep people building. Delivered inside the Promptathon and the AI Adoption Managed Program.
Operate
Models change monthly. Someone has to run this and keep proving it works.
QA
Agent Assurance
Agents re-tested as models change, with evidence formatted for internal audit and board reporting.
QW
AI Adoption Managed Program
Adoption telemetry, enablement clinics and quarterly outcome reviews.
QW
Fractional CAIO
Standing executive advisory from operators who have run technology inside banks, insurers and managed service providers.
QP
Data Platform Management
Proactive database management, consolidation and performance tuning for the estate your AI retrieves from.
QP
Platform AI Ops
Running the estate: monitoring, cost control, lifecycle and tuning. Delivered inside AI Platform Engineering.
QA
Application managed service
Deployed agents and applications run and monitored against a scorecard. Delivered inside Agent Assurance.
HONEST SCOPE
Who we are not for.
We are not the right firm if you want a demo rather than a decision, if there is no executive sponsor behind the work, or if the intent is to explore rather than act. We also do not chase Fortune 100 procurement cycles. At that scale you are better served by a firm built for it, and you will get a straighter answer from us saying so than from a proposal that pretends otherwise.
Which of these do I need first?
Start with readiness
You know AI matters. You cannot yet say where it pays off, or defend a plan to the board.
Start with the foundation
You know exactly what you want to build. The estate underneath it will not carry the weight.
Start with the work
One process, one team, one number. Prove it there before anything gets scaled.
Answered once, here, so no page has to repeat it.
Fixed-fee discovery
Every engagement opens with a discovery of published scope and duration, at a fixed fee. It ends in a decision, not a proposal to do more discovery.
A scoped build, capped
Build effort varies with process complexity and integration, so discovery scopes it first. From there we work to a not-to-exceed cap with weekly burn reporting. You always know what is left.
You own the output
Your deliverables and work product transfer to your team: designs, prompt libraries, standards, code and documentation. Our methods and evaluation framework stay ours, licensed to you for internal use.
We do not resell
You license your own platform and models directly. We take no commission or margin on what you buy. Where a vendor offers partner incentives, we disclose them before we recommend anything.