Independent AI consulting & implementation, now taking on new projectsStart a conversation
Services

Advice and delivery, from the same person

Two kinds of work, offered separately or together: consulting to decide what is worth doing, and implementation to build it and put it into use.

01 · AI consulting

Decide where AI fits, and where it doesn't

The goal is a short, specific list of opportunities worth pursuing, with a realistic view of effort, risk and benefit for each.

Useful if you're under pressure to “do something with AI” and want a grounded plan, if your team has tried tools without much to show for it, or if you're about to buy an AI product and want an independent view first.

Example prioritization from an assessment: each candidate task placed by expected value and ease of delivery. Illustrative.
  • Opportunity assessment

    Interviews with the people who do the work, a look at the tools and data involved, and an honest view of which tasks suit AI.

  • Roadmap & prioritization

    A sequenced plan that starts with a contained, measurable first project and builds on what is learned.

  • Build, buy, or both

    An independent comparison of off-the-shelf AI products, AI features already in your software, and custom builds.

  • Data readiness

    Where the information an AI system needs actually lives, how clean and reachable it is, and what has to change first.

  • AI usage policy

    Practical rules for how staff use AI tools: what data may go where, which uses need review, and who is accountable.

  • Team enablement

    Hands-on sessions that help your team use AI tools well, with examples from their own work.

What you receive

A written, prioritized set of recommendations

Specific enough to act on, and honest about what isn't worth doing.

  • Summary of the processes reviewed and the opportunities found
  • A ranked shortlist, with the reasoning behind each ranking
  • A recommended first project, scoped well enough to estimate
  • Data, access and risk issues to resolve before building
02 · Implementation

Design, build and deploy agents and workflows

Working systems connected to your tools and data, tested on your real examples, and handed over so your team can run them. Implementation can follow an assessment or start directly if you already know what you want built.

Agent description

Each ticket is classified, the customer's order and history are looked up, and a reply grounded in your policies is drafted for an agent to check and send.

AI workflow
  1. Ticket createdTrigger
  2. Classify topic and urgencyAI step
  3. Look up order and historyLookup
  4. Draft reply from policyAI step
  5. Agent reviews and sendsApproval
Draft reply · ticket 1182Waiting for agent review

Hi Sam, thanks for getting in touch. Your order left our warehouse on Tuesday, and the carrier's tracking currently shows delivery on Thursday. If it hasn't arrived by Friday, reply here and we'll arrange a replacement.

[1] Order record[2] Delivery policy §2

Example output, illustrative only

  • AI agents

    Agents that carry out multi-step tasks, such as reading inputs, looking things up and taking actions in your tools, within clearly defined permissions.

  • Workflow automation

    Event-driven workflows that combine ordinary automation with AI steps, on a platform that suits your team or as custom code.

  • Knowledge assistants

    Assistants that answer from your documents and data, cite where the answer came from, and say when they don't know.

  • Integrations

    Connections to email, CRM, helpdesk, accounting, document storage and databases, using official APIs wherever they exist.

  • Testing & evaluation

    A test set built from your real work, used to measure quality before launch and to check that later changes don't make things worse.

  • Support & improvement

    Optional monthly support after launch: monitoring, fixes and adjustments as your processes change.

What you receive

A working system you own

On tools and platforms: the right choice depends on what you already use, who will maintain it, and your data requirements. The recommendation and its reasons are always given to you in writing.

  • A working system, deployed in your environment or accounts
  • Test results against an agreed set of real examples
  • Documentation: how it works, how to run it, how to change it
  • Training for the people who will use and look after it
Ways to engage

Start small, then decide

Every engagement is scoped and quoted in writing after the first call. You know what you're getting, what it costs and the timeline before any work begins.

01 · Assessment

AI opportunity assessment

A defined review of your processes and a written, prioritized set of recommendations.

Pricing
Quoted in writing
Timeline
Set in the quote
02 · Project

Design & build

One agent or workflow, taken from design through testing to deployment and handover.

Pricing
Quoted in writing
Timeline
Set in the quote
03 · Ongoing

Support & advisory

Continued improvement and monitoring once systems are live, or a regular sounding board.

Billing
Monthly
Scope & price
In the written quote
Honest fit

When this probably isn't the right fit

Saying this up front saves everyone time.

  • You need a large team for a multi-year transformation program on a fixed deadline. A one-person practice is the wrong shape for that.
  • The process needs results that are always exactly right with no human check, and a mistake would be costly. Conventional software or a person is often the better choice.
  • You want AI mainly for the announcement rather than to solve a specific problem.
  • The underlying process isn't defined yet. It usually pays to settle how the work should be done before automating it.
Next step

Have a process you suspect AI could help with?

Start with a conversation about how the work is done today. You'll leave with an honest view of whether AI is worth pursuing for it.