Pick your context. The workflow routes itself to you.
The workflow adapts to who you are. Different capabilities surface depending on your context.
Practical automation and AI for growing businesses. No bloated enterprise playbooks.
I start with the workflows bleeding the most time: customer ops, reporting, lead handling, internal admin. One well-placed automation can reclaim 10+ hours a week. You get a full workflow audit document within the first week: every bottleneck mapped, every quick win scored by effort vs. impact. We pick the top three, build them in weeks, and the saved time funds the next round of improvements without touching your budget. Most clients see ROI before the first invoice is paid.
You don't need a data science department. I build lean AI systems: n8n automations, LLM-powered tools, smart internal apps that a small team can run and maintain. Every system comes with documentation your team can actually follow, a handover session, and a 30-day support window. I design for the person who inherits it, not the person who commissioned it. Right-sized technology, not oversized contracts.
You've probably seen AI strategies designed for companies ten times your size. I build phased roadmaps that match your budget, your team, and your growth pace. You walk away with a 90-day plan, a 12-month horizon, and clear decision points for when to invest more. Every phase has a measurable trigger, not a calendar date. So you invest in AI that compounds, not AI that gathers dust.
Most people aren't sure. That uncertainty is expensive. A 30-minute conversation will give you clarity on whether this is the right move.
Let's figure it out →The goal isn't a dependency on me. It's a team that knows what they're doing. I run hands-on sessions tailored to your actual workflows: how to prompt effectively, how to maintain your automations, how to spot the next opportunity. No slides about the history of AI. Just practical skills your people use the next morning. You get session recordings, reference guides, and a Slack channel for questions that come up later.
AI moves fast. What worked three months ago might not be the best approach today. I offer lightweight monthly retainers for SMEs who want ongoing optimisation: new workflow builds, model upgrades, troubleshooting, and a standing call to pressure-test ideas. No long contracts. Month-to-month. You keep it because it's valuable, not because you're locked in.
Here's what working with me actually gets you.
I've built and shipped multiple revenue-generating AI products. Not demos. Things people actually pay for. I'll take your idea, validate the technical approach in a day, pick the right stack, and get an MVP live in weeks. You get a working product, not a pitch deck about a product. At Zipline I helped scale operations 20x. I bring that same bias toward shipping to every founder engagement.
Your first workflow removes 5 hours a week. That unlocks capacity to build ten more. I design automation architectures: n8n, multi-agent orchestrations, LLM pipelines that create compounding returns, not one-off wins. Every automation is documented, version-controlled, and built to survive your next pivot. I hand over systems that your future engineer can extend on day one, not reverse-engineer for a month.
What to build, buy, or skip. I'm vendor-neutral. I'll tell you when the shiny new model is overkill and when a simple automation beats a complex AI system. You get a decision matrix: every AI component scored on cost, speed-to-ship, and switching risk. Founders need clarity, not more options. I've made these calls at every stage from pre-seed to scale-up, and I'll save you months of expensive experimentation.
Most people aren't sure. That uncertainty is expensive. A 30-minute conversation will give you clarity on whether this is the right move.
Let's figure it out →Most startup AI is built to demo. I build to scale. You get a technical architecture document that covers model selection, data pipeline design, cost modelling at 10x and 100x usage, and infrastructure choices that won't bankrupt you when traction hits. I've seen what breaks when startups grow fast. At Zipline, I watched systems go from prototype to operating at national scale. Your architecture should anticipate success, not just prove a concept.
AI isn't just your product. It's your unfair advantage in getting that product to market. I build the automations that let a team of three operate like a team of thirty: AI-powered lead scoring, automated outreach sequences, intelligent onboarding flows, and analytics pipelines that tell you what's actually working. You ship faster, learn faster, and spend less doing it.
Strategy, governance, and enablement. In that order.
Most enterprise AI strategies don't survive first contact with the board. I build roadmaps that are grounded, realistic, and framed around business outcomes, not technology for its own sake. You get a board-ready document with prioritised use cases, timelines, cost projections, and a vendor evaluation framework you can reuse. Each use case is mapped to a P&L line. No fluff, no filler. Just the strategic clarity your leadership team needs to commit.
I've built AI policies from the ground up, starting with what organisations care about, what risks are real vs theoretical, and how AI aligns with existing compliance frameworks. You get a complete policy suite: acceptable use guidelines, data handling protocols, model risk assessment templates, and an escalation matrix. Everything mapped to your existing compliance structure so legal doesn't start from zero. Governance that accelerates adoption, not blocks it.
Technology without adoption is expensive infrastructure. I design enterprise enablement programmes: from executive workshops that align leadership, to hands-on training for the teams using these systems daily. Deliverables include a tailored curriculum, session materials your L&D team can reuse, adoption scorecards, and a champion network that sustains momentum after I leave. The goal is self-sufficiency within one quarter.
Pilots fail when they're designed to impress, not to scale. I structure pilots with built-in graduation criteria: clear success metrics, a defined timeline, and a pre-planned path to production. Every pilot has a kill switch and a scale-up trigger. You know exactly what good looks like before you start, so the decision to go wider is based on evidence, not enthusiasm.
AI programmes stall when they can't prove value in the language the CFO speaks. I build KPI frameworks that track what actually matters: time saved, error reduction, cost avoided, revenue influenced. Every metric ties back to a business outcome your finance team recognises. You get a live dashboard spec and a quarterly reporting template that makes the case for continued investment automatic.
Most people aren't sure. That uncertainty is expensive. A 30-minute conversation will give you clarity on whether this is the right move.
Let's figure it out →I've deployed this in live public sector environments. I know what works.
As Head of Innovation at Digital Jersey, I've led sandbox projects taking emerging technology from concept to live public sector deployment. That means real-world testing with real stakeholders, not lab experiments. You get a sandbox framework document, a stakeholder engagement plan, risk registers tailored to public scrutiny, and a path from experiment to operational service. Structured enough for governance, agile enough to actually move.
The biggest mistake public sector organisations make is buying tools before building frameworks. I help establish the right AI governance structure first: ethical guidelines, accountability chains, data handling protocols, and impact assessment templates. These aren't academic exercises. They're practical documents your procurement team can attach to every RFP. So when you buy, you're buying the right thing with the right safeguards already in place.
Most people aren't sure. That uncertainty is expensive. A 30-minute conversation will give you clarity on whether this is the right move.
Let's figure it out →The goal is a team that carries things forward confidently, not dependency on an external consultant. I run training designed for public sector: clear, jargon-free, built around the constraints you actually operate in. Deliverables include role-specific training modules, a decision-maker briefing pack, and a self-assessment toolkit your team uses to evaluate future AI opportunities independently. I measure success by what happens after I leave, not while I'm there.
Public-facing AI has zero margin for error on trust. I design citizen-facing applications: intelligent service routing, automated case triage, all with transparency built into every interaction. Every system includes human escalation paths, plain-language explanations of how decisions are made, and accessibility standards baked in from day one. The public should understand what the AI does and why. No black boxes.
Scaling AI across government doesn't mean flipping a switch. I design phased rollout plans that start with one department, prove the model, then replicate with adjustments for each team's context. You get a rollout playbook with department-specific adaptation guides, shared service opportunities, and a governance structure that scales without creating bureaucracy. Each phase builds on the last. No department gets a copy-paste. They get a version that fits.
The unfiltered version. For people who want to build something together.
Since 2024 I've been building a startup alongside Digital Jersey. AI-native, built on tools I know work, solving a problem I've watched go unsolved for years. I'm doing the full thing: product, go-to-market, fundraising conversations, the lot. Being operator and founder simultaneously means I'm not advising from theory. Every recommendation I give clients, I'm pressure-testing in my own company the same week. That's the difference.
No vendor allegiances. Only preferred outcomes. My toolkit is whatever gets the job done: n8n for automation, Claude for intelligence, Lovable for rapid UI, Vercel for deployment, Gemini for multimodal work. I pick tools based on what ships fastest and scales best, not who's paying referral fees. I move fast, care about quality, and document everything so you're never locked into needing me. That last part matters more than people think.
Co-founder conversations, technical partnerships, associate engagements, interesting problems. If the problem is real and you're serious about solving it, I want to hear about it. I'm selective because capacity is finite. I'd rather do three things brilliantly than ten things adequately. Best engagements start with a real conversation, not a brief. If we're a fit, you'll know quickly.
Most people aren't sure. That uncertainty is expensive. A 30-minute conversation will give you clarity on whether this is the right move.
Let's figure it out →n8n is the backbone. Hundreds of production workflows running right now. Claude handles the intelligence layer for everything from content to code review to customer interactions. Lovable lets me build production-quality UIs in hours, not weeks. Vercel deploys everything. Supabase for data. Google Gemini when I need multimodal. This isn't a list of logos. It's the stack I use every day, and the one I'll build your systems on.
Sprint-based builds for when you need something shipped. Advisory retainers for when you need a thinking partner. Equity conversations for the right opportunity. I don't do day rates or time-and-materials busywork. Every engagement has a defined outcome, a timeline, and a clear definition of done. If it's not working, we stop. No hard feelings, no drawn-out wind-downs.
Built an AI-powered interior design tool for a small architecture practice. What used to take the owner days of manual mood-boarding and client prep now runs in under an hour. End to end.
Designed and deployed a multi-agent orchestration for a marketing agency. Raw call transcripts go in; structured briefs, social posts, blog drafts, and email sequences come out. No human handoffs.
Ran a full operational audit for an engineering firm. Mapped every process, identified automation opportunities, and delivered a phased roadmap with ROI projections the board signed off on in one session.
Built a platform that lets L&D teams create bespoke content paths for every employee. Instead of one-size-fits-all training, each team member gets a tailored learning journey matched to their role and skill gaps.
Systems-level thinking and structured problem-solving. The analytical foundation behind every automation architecture and AI strategy.
Grew delivery capacity 20x across West Africa. Won multi-million dollar government contract in Ivory Coast. Optimised autonomous flight operations at continental scale.
Built B2B product suite from zero. Led commercial strategy through growth phase to successful acquisition in 2023. Revenue-focused product development.
Designed and launched startup incubator and accelerator programmes. Led sandbox projects deploying AI, IoT, and drone technology in live public sector environments.
AI-native product built on proven tools. Dual perspective as operator and founder makes every advisory engagement sharper and more grounded.
78 tools, 6 skill domains, 5 organisations, and the processes that tie them together, mapped as a force-directed knowledge graph. Tap any node to explore its connections.
I'm an AI trained on Seb's background. Ask about his work, approach, or whether he's the right fit for what you're building.
You've seen the process. The proof. The toolkit.
Now let's run it on your business.