Field notes from production AI.
Practical writing on AI integration — architecture choices, deployment patterns, security defences, and what real B2B AI costs to build and run.
- Pricing3 Jun 20269 min read
What does a custom AI agent actually cost? An honest breakdown
A line-by-line breakdown of the real cost of designing, building, and running a custom AI agent for B2B — engineering, model spend, observability, and the operational tax that's missing from every vendor pitch.
Read article - Security27 May 202610 min read
How to prevent prompt injection: defense patterns that actually work
Prompt injection isn't going away. A practical look at the defence patterns that hold up in production B2B systems — instruction firewalls, retrieval isolation, output validation, and the limits of each.
Read article - Production20 May 202610 min read
How to deploy AI in production safely: a B2B leader's checklist
The production-readiness checklist for AI systems that handle real customer impact. Covers evals, output validation, observability, security review, and the failure modes nobody warns you about.
Read article - Adoption13 May 202611 min read
How to integrate AI into your existing business in 90 days
A 90-day, three-phase playbook for B2B operators integrating AI into a real business — without burning a year on pilots that never reach production.
Read article - Architecture6 May 20269 min read
Self-hosted LLMs vs the OpenAI API: when each makes sense for B2B
A practical decision framework for B2B leaders weighing self-hosted open-weight models against managed LLM APIs — covering compliance, latency, cost at scale, and the operational tax that's rarely on the slide.
Read article
Want this kind of thinking applied to your project?
Every post here came out of real client work. Tell us what you're trying to do — we'll tell you honestly whether we can help.
No deck. No drip campaign. One reply.