Vol. I — fbmhdl Dispatches London · May 2026 · No. 01

On the practice of building
conversational AI for businesses.

A long-form note on what we do at fbmhdl, how we do it, and the assumptions we are happy to argue with. We build LLM-powered assistants for support, internal operations, and regulated industries.

i. Practice

A note on how we work.

Most conversational AI today is bolted-on no-code chained to a vendor's platform. It looks impressive in a sales deck and falls apart the first time a customer asks something the demo didn't cover.

We do the opposite. We write actual software in your stack, with your data, your guardrails, and your engineers in the room. The repository is yours from day one. If we walked away tomorrow, the assistant would keep running.

We work in two-to-twelve week engagements. We do not sell licences, recurring "AI seats," or anything that looks like a SaaS tier. You pay us to build something, then you own it.

This is unfashionable. The market wants platforms; we want code. But for a serious deployment, in a regulated industry, with the audit and data-handling constraints that come with it, code wins.

ii. Work

What we build, and what we have built.

01
Customer-facing

Assistants for support

Account questions, order tracking, the long tail of "talk to a human." Web, WhatsApp, Messenger, Telegram, voice, or your own mobile apps.

02
Internal

Copilots for ops

Domain-trained agents for ops, sales, and back-office teams. Connected to your CRM, ticketing, knowledge base, internal APIs. Audit trail and RBAC by default.

03
Integration

The middle layer

Telephony, payments, identity, evals, logging. We do this on its own for teams that have built the assistant but can't get it into production.

Logistics · courier Live

ABC Parcel

Tracks parcels, reschedules deliveries, changes addresses, handles shipping FAQs. Built on patterns we'd use for a national operator.

abc.fbmhdl.com →
Retail banking Live

Banking assistant

Account-aware support, onboarding, scoped tool calls, full PII scrubbing and audit log for every interaction.

demo.fbmhdl.com →
iii. Process

Three phases. No retainer trap.

01
Week 1 · Discovery

One week, fixed fee

We map the workflow you want to automate, the data the assistant needs, and what "good" looks like. Output is a written scope and a working prototype.

02
Weeks 2-8 · Build

Iterate on a real surface

We build against your real channels and real integrations. Weekly reviews, staging environment from day one, eval suite so changes don't quietly break.

03
Hand-over

You run it

Code in your repository, docs in your wiki. A short handover with your engineers. We're available afterwards on retainer or by the hour, but you don't need us to be.

iv. Correspondence

Write to us.

A few lines about what you're building is enough to start. We reply within one working day. If we're not the right fit, we'll say so.

Phone / WhatsApp
+44 7783 990485
Office
182-184 High Street North, London E6 2JA