Most "AI automation agencies" are three-person shops in matching hoodies pretending to be McKinsey. I'm one senior operator with 16 years of process discipline and a working agent stack of my own — used in anger, every day. The result is automation that actually replaces hours, not another SaaS line on your invoice.
Plenty of AI automation agencies will sell you a 60-slide AI strategy. I'd rather build something useful in week one. These are the engagements I'm taking on right now.
A 2-week paid engagement. I shadow the workflows you suspect could be automated, score them on time saved vs. complexity, and ship a prioritised backlog with prototypes for the top three. You walk away with a real plan — even if you don't engage further.
The build phase. I implement the automation — research agents, content pipelines, reporting bots, internal copilots — using the LLMs and platforms that fit. You get working software, documentation and a hand-off your in-house team can maintain.
For teams that don't want to maintain their own agents. I keep your automations alive, monitor failure modes, and ship iterations as your needs evolve. Capped hours, monthly reporting, no auto-renewal lock-ins.
For teams who want to own the work themselves. Half-day or full-day sessions covering agent design, prompt engineering, automation hygiene, and how to spot the workflows worth automating. Hands-on, not theory.
No multi-month onboarding. No fifteen kick-off calls. Discovery starts week one; something is shipping inside a fortnight.
I sit with the team. Read the docs. Watch the work happen. We score opportunities by hours saved and confidence — and decide what's actually worth automating versus what just looks impressive in a slide.
Choose the right model, the right orchestration, the right integration points. Sketch the agent's inputs, outputs and failure modes. Sign off on scope before a single line of code is shipped.
Build, test, integrate, document. The automation runs in production — not in a demo. Your team can see it work, audit its decisions, and call it themselves.
Models change, processes change. Light-touch ongoing support keeps the automations honest and your team's roadmap moving — without locking you into a fat retainer.
I run my own business on this stack, which is a useful tell. If I wouldn't put it in production for myself, I wouldn't sell it to you.
Multi-agent chains that take a topic brief, run deep research, produce a long-form draft and hero imagery, and deliver into the team's working tools. The kind of system that turns a content backlog into a steady drumbeat.
Fully autonomous pipelines that run on a schedule, pull data, run analysis, and ship the output into the channels your team actually checks. Real, production-grade — not a demo dashboard.
A repeatable methodology that runs a brand and its competitors through GPT, Claude, Gemini and Perplexity, scores entity accuracy and citation share, and outputs a measurable report month over month.
For teams drowning in Notion or Drive. A retrieval-augmented agent over your internal docs, channels and meeting notes — so "what did we decide about X?" becomes a one-line query, not an archaeology dig.
A scheduled agent that watches a portfolio of competitor sites, social profiles and SERPs, flags meaningful changes, and produces a weekly digest. Replaces an analyst doing the same thing every Monday morning.
Monthly client reports that pull from analytics, search consoles, paid platforms and LLM audits — assembled into a narrative document with no junior touching a spreadsheet at midnight.
Not because solo is universally better — but for the kind of project most teams actually need, the agency model is overhead you'd rather not pay for.
| Typical AI automation agency | Dog on the Table | |
|---|---|---|
| Who actually does the work | Mostly junior implementers, supervised by a senior consultant on calls only | A 16-year senior operator. Every line of the build. |
| Time to first shipped automation | 6–10 weeks (after onboarding, kickoffs, alignment sessions) | ~2 weeks from discovery to first working automation |
| Platform commitment | Locked into the agency's preferred stack (often Zapier/Make-only) | Platform-agnostic — n8n, LangGraph, custom code, or whatever fits |
| Account managers between you and the work | 2–3 (PM, CSM, sometimes a strategist) | Zero. You email me. I reply. |
| Demonstrated working agent stack | Usually case studies in slides only | A production agent stack I run on my own business — used in anger |
| Lock-in / minimum commitment | 6–12 month retainers common | Per-project. Optional light retainer with no auto-renewal. |
The useful work is designing and building autonomous or semi-autonomous workflows that take repetitive operational, content or research tasks off your team's plate. That includes agent chains (research → analysis → output), automated reporting pipelines, content production systems, and integrations between the tools you already use. Anything else is just consulting with an AI sticker on it.
Traditional automation consultants tend to live inside one platform (Zapier, Make, n8n) and chain triggers. I work platform-agnostic and lead with AI-native agents — workflows that read documents, summarise meetings, run scheduled research, or produce drafts. The output is autonomous, not just a sequence of triggers.
Often the right answer is a mix: agents for the cognitive parts, classical automation for the plumbing. The interesting work is deciding which is which.
No. Most engagements start with a discovery session where I learn how your team actually works, then I prototype something concrete within a couple of weeks. You see real output before signing off on anything more ambitious. The technical work is mine; the domain knowledge is yours.
Most engagements start with a fixed-fee automation audit (a couple of weeks, prototypes included) so you know what you're buying. From there we scope a build — usually project-based — followed by an optional light retainer if you want ongoing maintenance and iteration. No 12-month contracts.
Yes — Brighton. I work with UK, European and US clients remotely. For UK and EU engagements I can travel for kickoffs and stakeholder sessions where it matters.
Even better. I'll often act as the senior specialist your team doesn't have a full-time hire for — designing the agents, training the team to maintain them, and stepping back. The aim is to leave you self-sufficient, not dependent.
If you've been told you need "an AI strategy" but what you actually want is something working in production by month-end — let's talk. Most engagements start with a paid audit.