Save your team 15+ hours a week with AI that actually works.
Most AI projects burn six months and ship nothing. Mine scope to a measurable outcome — typically 15–30 hours of weekly time recovered within 8 weeks, paying back the engagement inside one quarter. Or I tell you in 30 minutes when AI isn't the right answer.
Sectors served
- B2B SaaS
- Professional services
- E-commerce
- Manufacturing
- Education
- Local services
- Marketplaces
- Agencies
What I do
Each service scoped to a measurable outcome and delivered in 4–8 weeks. Fixed price, agreed before kickoff.
GPT & Claude Integration
Cut response and triage times 70–90% by wiring GPT-4 or Claude into the tools you already use. Production-ready in 4–6 weeks. Token costs capped from day one.
AI Strategy & Roadmap
Spend 2 weeks getting a 90-day AI roadmap with ranked ROI estimates — and avoid the typical 3–6 months and £40k spent building the wrong thing first.
Custom AI Workflow Builds
Production-grade AI workflows shipped in 4–8 weeks — monitored, documented, and operable by your team without me. Fixed price, no surprise bills.
AI Content & Knowledge Systems
Answer 80%+ of repeat queries automatically with AI grounded in YOUR content — not the open internet. Cut support volume + onboarding time without a hire.
AI for Customer Operations
Triage and route inbound in seconds, not days. AI that understands context — typical results: 70%+ less manual triage, MQL→SQL conversion up 20–35%.
Data Pipelines & AI-Ready Reporting
Get back 4+ days every month. Self-updating reports across your stack, queryable by AI in plain English. Anomalies surface in hours, not at month-end.
How I work
Three phases, scoped to outcomes not hours. Most engagements run 4–8 weeks — short enough to ship, long enough to do it properly.
Discover
Two-week audit of where the team's time actually goes. We surface the manual work, score it for impact and effort, agree what's worth automating.
Design
Sequenced roadmap with the highest-impact wins first. Tooling chosen to fit your stack (n8n, Make, GPT, custom). You sign off on what we ship, in what order.
Build & Hand-off
Build the automations, run them through real data, document them. Handover means your team owns and can change them — not a black box that needs me forever.
Common questions
Things teams usually want to know before scoping a project.
How long does a typical engagement take?
Most projects run 4–8 weeks. A discovery / roadmap engagement is closer to 2 weeks. Builds scope to outcome, not hours — we agree what 'done' looks like before we start.Which AI models and tools do you typically use?
I work primarily with GPT-4 and Anthropic Claude, plus open-source models where they fit (Mistral, Llama variants). On the integration side: APIs into your existing CRM, helpdesk, CMS, or product. For data: BigQuery, Postgres, vector DBs (pgvector, Pinecone). Tooling follows the problem, not the other way round.How do you charge?
Fixed price per project, scoped to outcome. You know the cost before we start. Long retainers are rarely the right shape for AI work — most engagements are project-based with optional ongoing support and evals after handover.Do you work with both enterprises and smaller businesses?
Yes. Global enterprises with security, compliance, and audit requirements; local businesses with one operational question worth answering. The technical work is similar — what changes is the scope, the governance, and how careful we need to be with the data.Will my team be able to maintain what you build?
Every project includes documentation and handover. You shouldn't need me to keep things running — that's the test of a good build. I'm available for ongoing support, evals, and tuning if you want it, but never as a dependency.What if AI isn't the right answer?
I'll tell you. Most business problems aren't AI problems — they're process, tool-stack, or data-cleanliness problems. The work might be a clean integration, a better workflow, or removing a tool that's getting in the way. Not every engagement involves an LLM, and that's fine.
What you get
Working with me is meant to feel low-risk. The hard part is the build — everything else is fixed up front.
Fixed price, agreed first
Scope and price agreed before kickoff. No surprise bills, no scope creep, no per-hour invoicing.
Free 30-min scoping call
You leave with at least one identified opportunity (or an honest 'AI isn't your answer here'). No commitment, no slide deck.
Your team owns it day one
Every project includes docs, hand-over, and training. I'm not a dependency — I'm a sprint.
Outcome-scoped, not hours
Every engagement has a concrete deliverable you can point at and measure. If we can't define it, we shouldn't be doing it.
Get 15+ hours back every week — payback in one quarter.
Free 30-min scoping call. By the end you'll have at least one specific AI opportunity in your business with an estimated time-save and cost — or I'll tell you why AI isn't the right answer for the problem (and what is).