Insights
Lessons from driving business value with AI.
Everything here comes out of something we built, run or delivered for a client — including the parts that went wrong.
Agentic AI
Letting the model choose, with the tools the series built
Handing a Bedrock model the order gateway and the Code Interpreter sandbox as tools, with AgentCore Memory as context, and letting it decide what to call for a question nobody wrote code for, while…
4 Oct 2026
Read the pieceAI Value Diagnostics
Harness engineering, applied to an infographic
What a harness is, when a plain prompt is enough, and why an infographic needs one. Then McKinsey's State of AI 2026 on one A1 sheet, built under a harness and free to download.
30 Sep 2026
Read the pieceAI Value Diagnostics
A year of record AI spending. The value needle didn't move.
McKinsey's State of AI 2026 leaves the value figures where they were, 37% attributing any profit impact and 6% getting significant value. What the 6% do differently is not technology.
28 Sep 2026
Read the pieceAI Value Diagnostics
AgentCore in the wild: where agents pay back, and what separates the 31% from the rest
Agents pay back when they are carefully placed: what the production successes share, the three disciplines that separate the 31% from the rest, and the buyer's side of the hands-on AgentCore series.
21 Sep 2026
Read the pieceAgentic AI
Knowing who your agent is acting for, with AgentCore Identity
Both halves of AgentCore Identity for one agent, inbound auth on the runtime and an on-behalf-of exchange brokered by a workload identity, credential provider and token vault, for a five-minute…
11 Sep 2026
Read the pieceAI Value Diagnostics
AI-first, humans in control: a deeper, faster, wider CMS assessment
How Levantar helped Highfield, one of the UK's top five awarding organisations, decide the future of its content platform — working AI-first, with people owning every judgement.
25 Aug 2026
Read the pieceAgentic AI
Letting an agent run code, without letting it run loose
Giving an agent a managed Python sandbox with AgentCore Code Interpreter, putting a file in, running code over it, and seeing what the session can and cannot reach.
21 Aug 2026
Read the pieceAI Value Diagnostics
The pruning has started. Most of it is cutting the wrong branch.
AI budgets are being cut on cost because value is the harder number. What the 2026 cost-governance data shows, and what to prune on instead.
19 Aug 2026
Read the pieceAgentic AI
Giving your agent memory that survives the session
Adding AgentCore Memory to an agent so it keeps conversation events within a session and recalls extracted user preferences across sessions.
14 Aug 2026
Read the pieceAgentic AI
An agent at the centre
Six best-of-breed apps that don't talk, four small open-source MCP servers, and an agent in the middle. What we built, what broke, and what it taught us.
6 Aug 2026
Read the pieceAgentic AI
Giving your agent tools with AgentCore Gateway
Putting AgentCore Gateway in front of an existing Lambda so an agent can call it as an MCP tool, authenticated from the first deployment.
5 Aug 2026
Read the pieceAI Value Diagnostics
How and why we built Upcast.Social
Why we built our own social tool rather than hiring or subscribing, how we validated it in Opptora first, and how it became a product with a reseller programme.
4 Aug 2026
Read the pieceAgentic AI
What AgentCore actually is, and getting a first agent running on it
How to deploy an agent container to Amazon Bedrock AgentCore Runtime with Terraform, invoke it, and see what the platform actually gives you.
24 Jul 2026
Read the pieceAI Value Diagnostics
Why we built Opptora first
We built our own validation platform so we would only build what we could prove had value. It now sits underneath the assessment we run for clients.
23 Jul 2026
Read the pieceAgentic AI
Claude + ElevenLabs → narrated MP4, in four languages
The narrated walkthrough, localised — Claude translates, ElevenLabs voices it in four languages, ffmpeg stitches each one.
29 May 2026
Read the pieceAgentic AI
Claude + Playwright + AWS Polly → narrated MP4
The same flow with video output: Claude, Playwright and AWS Polly producing a narrated MP4.
26 May 2026
Read the pieceAgentic AI
Claude + Playwright → interactive walkthrough
Point Claude at a feature and get an interactive walkthrough out the other end.
16 May 2026
Read the pieceAI & LLM Security
Building a multi-agent security pipeline with Google's A2A protocol and GitGuardian
Wiring multiple agents together with Google's Agent Development Kit and the A2A protocol, with GitGuardian scanning content for secrets along the way.
19 Dec 2024
Read the pieceContact
Talk to us before your next AI investment.
If something here matches a problem you have, the first conversation is diagnostic, not a pitch.