Augment your teams and processes.

Delivering with teams, not for them.

01The thesis

Most AI projects fail because organisations bolt on tools before they understand where the value is. I start with people and process, then build the intelligent systems that compound their effort.

02Practice
  • Find the value
  • Your Claude, dialled in
  • Agents you can trust
  • Memory that scales
Book a call
Engagements open · Q3 2026 · two slots

Agentic Development Lifecycle.

Memory and context

Hybrid stores plus retrieval engineering decide whether agentic systems compound value or degrade over time.

Eval-driven development

Evals before code. Agent acceptance criteria validated in CI before production traffic.

Uncertainty quantification

Uncertainty quantification is the discipline that matters most in regulated environments: having visibility of your systems' determinism.

Agentic engineering

AI can look great in a demo.

Agentic systems, much like software nowadays, are easier to demo than to deliver. Memory, context, and observability around uncertainty decide whether they survive renewal.

Memory architectures decide whether the system compounds over time or has a tendency to degrade. Recursive evaluations, self-learning, and uncertainty quantification keep it on the compounding side.

I run Dan's week. And on Sunday nights, we improve me.

·adminOS

I deliver functionality. For others, and for the system I run inside.

·forgeOS

Technologies I am well acquainted with.

Stacks · models · platforms
OpenAI
Anthropic
Python
TypeScript
Node.js
React
LangChain
LangGraph
MCP
AWS
Azure
Google Cloud
OpenAI
Anthropic
Python
TypeScript
Node.js
React
LangChain
LangGraph
MCP
AWS
Azure
Google Cloud
Bedrock
Docker
Cursor
n8n
Salesforce
HubSpot
ServiceNow
Power Automate
Vercel
GitLab
Git
Postman
Bedrock
Docker
Cursor
n8n
Salesforce
HubSpot
ServiceNow
Power Automate
Vercel
GitLab
Git
Postman

"Chase determinism where the system allows it. Quantify uncertainty where it doesn't."

Dan Charles · Founder's note
Dan Charles, founder
Dan Charles · Founder

Photo · 2025

About Dan

Collaborative Solutions was founded by Dan Charles to empower individuals and organisations to thrive in the evolving landscape of modern work, leveraging cutting-edge technologies like temporal memory to contribute to meaningful change.

Eight years across financial services, energy, and telecoms, in multifaceted roles spanning pre-sales, post-sales, and product development from ideation through deployment. The kind of regulated, high-stakes work where systems have to be both performant and auditable.