Fractional product leadership for B2B SaaS teams putting AI in the product. I help you work out what is worth building, build it alongside your team, and measure whether it moved the numbers. Fifteen years in, I still build.
An AI feature that works in the pilot and never reached real users, evals or a cost line.
The honest answer is a list of experiments. Nobody owns turning it into a roadmap with dates.
Quality, usage and spend are not measured, so the next decision is a guess.
Chatbots and autocomplete, not agents on the codebase. Delivery is slower than it needs to be.
Fractional or interim head of AI and product. I decide which use cases are worth building, set the roadmap and the cadence, own the evals, guardrails and governance, and hire the person who replaces me. Analytics and integrations come with it: the instrumentation to know what worked, and the patterns that turn a three-week integration into a five-day one.
A working proof of concept in days, not quarters. I build it myself with AI-assisted tools, on your data where possible, so the board, a customer or the team can react to something real before you commit a roadmap to it. Tariffik, a live commodity code lookup that helps businesses find the right tariff codes for customs declarations, is one I built and run this way, and there are dozens more behind it.
An AI feature taken from idea to production in four to six weeks, with your engineers. Use case chosen on evidence, build or buy decided, evals, guardrails, an audit trail and a cost line before launch, usage measured after. I have shipped chatbots, transcription and screening to 900 clients on a platform every product team could use.
Hackathons and hands-on Claude workshops on your own codebase: from chatbots to agentic development, skills that encode your standards, and real features at demo by five o'clock. I run these publicly as the Good Vibes series and privately for teams. You leave with working habits, not slides.
Plenty of AI consultancies build without product judgment. Plenty of product leaders have never shipped AI. I have spent the last few years doing both inside a scaled B2B SaaS: working out what was worth building, shipping it with the team, and measuring what actually moved.
Results from in-house roles, shown to demonstrate operating depth. They are not consulting-client outcomes.
Candidate chatbots, interview transcription and intelligent screening, shipped to production on a dedicated AI service and run as a platform for every product team, with the evals, usage measurement and governance documentation to match.
Global HRTech company, 900+ clientsTypical integration build, down from three weeks, after building an integrations team from zero to four and re-architecting the patterns. Net revenue retention moved with it.
Global HRTech companyEmployee number and first product hire. Set up the product function through the first external funding round and the growth to 90 people, then led AI and innovation across every product team.
Global HRTech companyFetch.je went from idea to launch in five days during lockdown, facilitated £250,000 of sales for 130 local businesses in three months, and was sold to Jersey Post.
Fetch.je, founderTwo weeks inside the product, the data, the AI work so far and the team. You get a written view of what is worth building, what to do first, and what it would take. Fixed fee, credited against anything that follows.
Put the measurements in place first: product analytics for the funnel, evals and usage for anything AI. Every decision after this is made on evidence rather than opinion.
Run the roadmap with your team, week by week, and build alongside them when that is the fastest route, until the changes are live and the numbers have moved.
Leave behind a product function, the instrumentation and the written playbooks that keep doing this without me.
For a founder or product lead who wants a second brain on AI and product, not a delivery team. Two calls a month and async access in between. You bring the decisions, I help you make better ones.
Embedded two or three days a week as head of AI and product. I own the roadmap and the cadence with your team, attend the standups, and ship.
A piece of work with a date: the two-week diagnosis, a proof of concept in days, an AI feature taken to production, or a workshop day with your team.
Dave Edwards has spent fifteen years in product management, analytics, growth marketing and engineering leadership, across B2B SaaS, e-commerce and marketplaces. He was the first product hire at a venture-backed global HRTech company and led product, then integrations, then AI and innovation through its growth from 18 people to 90 and 900 clients. Before that he ran DWE Digital as an e-commerce and growth consultancy, founded and sold Fetch.je, and ran a nine-person team with P&L at iProspect for adidas, Burberry, Coca-Cola and ASOS. He co-founded the Channel Islands Python User Group, mentors at Digital Jersey, and runs Good Vibes, a series of hands-on AI coding hackathons and workshops.
Tell me what is not growing. If I can help, I will say how. If I cannot, I will say who can. No forms, no sales call.
dave@dwedigital.com