SHORT ANSWER
Build an AI MVP by choosing one task where being right 80% of the time is still useful, wrapping a hosted model rather than training your own, and designing the interface around correction — users must be able to see, edit and accept output. Set an evaluation set before launch, measure task success and human-override rate, and only consider fine-tuning once the workflow itself is proven.
Answered by Andrew Crossley, Fractional Chief Product Officer · Updated 2026-08-01
AI products fail at the workflow layer, not the model layer. The model is a commodity; the useful part is where it sits in someone's job.
Without evaluation, teams ship confidently wrong output and discover the problem through churn instead of through data.
Drafting, summarising, classifying, triaging. Avoid tasks where a single wrong answer is unacceptable until you have evaluation and guardrails.
Use a gateway to a frontier model. Model choice is a tuning decision later; the workflow is the risk now.
Fifty to a hundred real examples with expected outputs. Run it on every prompt or model change. This is the single highest-leverage practice in AI product work.
Show the reasoning or source, allow editing, and capture every correction. Corrections are your training data and your quality metric.
Task success rate, human-override rate, time saved per task, and cost per successful task — not tokens processed.
Cheaper models, caching, retrieval or fine-tuning, chosen against the evaluation set rather than intuition.
FROM EXPERIENCE
The AI features that hold up in production are the boring ones: triaging inbound requests, drafting a response someone edits, extracting structured data from messy input. Each replaces a repeated manual step with a reviewable draft.
The ones that fail are open-ended assistants bolted onto a product, where the user has to invent the use case themselves.
Almost never at MVP stage. Hosted frontier models plus good prompting and retrieval cover the vast majority of use cases.
Ground answers in your own data, show sources, and design the UI so output is reviewed rather than auto-applied.
Model spend usually lands between £0.01 and £0.30 per task depending on model and context size. Measure cost per successful task, not per call.
IN SHORT
Five days to build your own AI-powered MVP, no engineering team.
Read moreTurning LLMs, RAG and agents into products that hold up in production.
Read moreAI-accelerated MVP delivery, scoped to one journey that earns.
Read moreThe self-paced course for founders shipping their first AI product.
Build your AI MVP in 5 daysTHE FRAMEWORK
MORE ANSWERS
Build an MVP by picking one job a specific user will pay to have done, designing the single shortest flow that completes it, and shipping only that. Validate demand before building — interviews plus a paid or committed signal. Build in four to six weeks with off-the-shelf infrastructure, launch to a narrow group, and measure whether they complete the job and come back.
A focused software MVP costs roughly £5,000–£15,000 built with AI-assisted tooling and a small scope, £20,000–£50,000 with a freelance or agency team, and £60,000–£120,000 for a complex or regulated product. The main cost driver is scope, not day rate: every extra flow adds design, build, testing and support cost. Budget 20% for post-launch iteration — that is where the learning is.
A focused MVP goes from idea to launched product in six to twelve weeks: one to two weeks of discovery and validation, four to six weeks of build, one to two weeks of pilot and fixes. Complex, integrated or regulated products take four to nine months. Delays almost always come from scope changes and slow decisions, not from engineering speed.