AI Agents
Agents that plan, call your tools, and finish multi-step work, with a person in the loop where it matters. For support, research, operations, and back-office work with clear steps and clear stop conditions.
Who this is for
- A support queue where most tickets follow a known path through your systems.
- A research or reporting task someone assembles from several sources every week.
- Operations work that means reading, deciding, and updating records in three tools.
- A team that wants to try agents without handing them the keys to production.
What is included. Tick what you need.
Artefacts, not adjectives. Each is something you can point to at the end. Tick the ones your project needs and send the list with your enquiry; we reply with a written scope.
Technologies we use for this
The relevant slice of our technology matrix. Nothing here that we cannot staff today.
- Python (Django, FastAPI, Flask)
- Node.js (Express, NestJS)
- PHP (Laravel)
- Go
- REST
- GraphQL
- gRPC
- WebSockets
- Celery
- Redis Queue
How we deliver it
The five steps every engagement goes through, in the form they take for this service.
Scope
A call with an engineer, then a written scope: what is in, what is out, and what it costs.
Architecture
Data model, API contract, and infrastructure plan, approved before code is written.
Build
The agent runs against real tasks in a sandbox first, with weekly evaluation results and a growing set of approved actions.
Harden
Tests on the paths that matter, error tracking, a performance pass, and a security review.
Launch and hand over
Production deployment, monitoring, documentation, and every repository transferred to you.
From the blog
- LLM security: prompt injection, data leakage, and what actually stops themMost LLM security advice is 'write a better system prompt.' The actual defenses are architectural: what the model can see, what it can do, and what never returns to a user unchecked.
- LLM observability: what to log, and what it actually catchesTraditional APM tells you a request was slow. LLM observability has to answer a different question: was the answer right, and why did the model say that?
- How to build an AI analytics assistant'Ask your data a question' is a text-to-SQL problem wearing a chat interface. The architecture that makes it trustworthy: schema grounding, guardrails, and a query the user can check.
Questions we get asked
What can an agent safely be allowed to do?
Read widely, write narrowly. Agents start with read access and reversible actions, and each irreversible action goes through a person until the evaluation results justify removing that step.
Which models do you use?
Anthropic and OpenAI models through their APIs, or open-weight models in your cloud when privacy or cost demands it. The agent is built so the model can be swapped.
How do you stop it going wrong?
Stop conditions, step limits, permission scopes, an audit log, and evaluation on every change. Failures are visible in a dashboard, not discovered by customers.
How is this different from automation?
Automation follows a fixed path. An agent decides the path from the goal and the tools. Many problems only need automation, and we say so when that is true.
What does it cost to run?
Cost is tracked per task type from the first week. Most agents settle well under the cost of the manual work they replace, and the numbers are yours to check.
Tell us what you are building.
We reply within one business day with how we would build it, what it would cost, and which engagement model fits.
- 01Tell us what you are building
A short form or an email. No deck required, and "not sure yet" is a fine answer.
- 02A call with an engineer
Within one business day. Technical questions get technical answers, from the person who would build it.
- 03A written scope and quote
Fixed price where the scope is defined. The document is yours whether or not you go ahead.