Fixed Scope

Pilot Build

A fixed-fee project to build and ship a working AI prototype.

Ideal for research teams, professional services units, founders, and businesses that need something working to test with users, demonstrate value, or validate an idea before committing to a full build.

The Problem

You have a strong idea for an AI-supported product or internal tool, but:

• No technical co-founder or in-house development capacity
• Limited budget to hire a full dev team
• You need something working to show users, funders, or management
• Previous experiments have stalled in "proof of concept" phase without real handover

You need a working prototype with clear ownership, documentation, and IP transfer, not another demo that dies when the consultant leaves.

What I Deliver

A focused pilot or supervised build when a working prototype is the fastest way to test a workflow change. The objective is not a demo, but a usable system with clear owners, governance, and handover so your team can run it without ongoing dependency.

Working Prototype

Not a slide deck or specification. A functional system you can use with real data and real users.

Documentation and Handover

Code walkthrough, prompt logic, architecture decisions, and failure modes documented. You understand how it works and where the risks are.

Full IP Ownership

All code, prompts, and documentation transfer to you. No ongoing licensing fees or dependency on proprietary systems.

Clear Success Criteria

We define upfront what "working" means. Not vague promises, but measurable outcomes (e.g., 80% accuracy on tagging, processes 100 records/hour, etc.).

How It Works

Days 1-2: Scoping and Success Criteria

Define narrow pilot scope: what exactly does this need to do? Establish success criteria (e.g., 80% accuracy on X, processes Y records per hour). Identify data sources and integration requirements.

Days 3-7: Build and Test

I build the prototype using AI-assisted development tools and a modern web or data stack, depending on requirements. Key activities:

Core functionality implementation
Integration with your existing tools/data sources
Testing on representative sample data
Documentation of failure modes and edge cases

Handover and Training

I walk your team through the code structure, prompt logic, and cost monitoring. You receive full documentation and IP ownership. If you have technical staff, they can extend it. If not, you can run it as-is or hire someone later to modify it.

Outputs

Working prototype aligned to a defined workflow
Handover pack (setup notes, roles/ownership, and operating instructions)
Risk and governance notes (controls required before broader rollout)
Pilot scorecard (success criteria for adoption, quality, and risk)

Investment

£7,500–£20,000 (fixed fee once scoped)

What moves the price:

• System integrations: Simple spreadsheet workflow (lower end) vs. API integrations with multiple systems (higher end)
• Data handling requirements: Public data vs. sensitive data requiring security/compliance measures
• Success criteria complexity: Basic proof-of-concept vs. production-ready accuracy targets
• Timeline constraints: Standard 1-week build vs. expedited delivery

Timeline: 3–7 days to working prototype (from scoping to handover)

Implementation Pilot (4-6 weeks)

A controlled rollout to embed the workflow change: governance cadence, stakeholder adoption, evaluation, and iterative improvement, ending with a maintainable operating routine.

Outputs

Implementation plan (owners, timeline, governance cadence)
Adoption package (training notes, documentation, internal comms draft)
Evaluation results against the pilot scorecard
Transition plan to steady-state operations (maintenance ownership and review cycle)

See the full playbook

Read the rapid AI prototyping guide for more detail on this approach.

Ready to build a pilot?

We start with a Discovery & Governance Sprint to scope the pilot and establish success criteria.

Apply for a Discovery & Governance Sprint

Questions? Email [email protected]