Built by someone who's done the work
The Praxis Agents Platform wasn't built on theory. It came from years of building AI agents for real businesses, and learning what actually works.
My education background is deeply technical, but I started my career in digital advertising - working with advertisers, agencies and technology companies. That background gave me a deep understanding of how marketing operations actually run: the tools, the workflows, and the places where things fall through the cracks.
In November 2023, when OpenAI released the Assistants API, I saw the shift immediately. Giving large language models the power to take actions - not just generate text - changed everything. I built my first agent from scratch that day.
Since then, I've built hundreds of AI agents for different businesses, using every major framework along the way: LangChain/LangGraph, OpenAI, Anthropic SDK, Google ADK and more. Every project taught me something new about what makes agents actually useful vs just a clever looking demo.
Praxis Agents is everything I've learned over the last 3 years, built into a single platform. Agents that remember your context, follow your rules, connect to your real tools, and always keep you in control. No black boxes. No surprises.
Why Praxis exists
Most AI tools give you access to a standard LLM and call it done. But the teams I worked with didn't need another place to ask questions - they needed something that could actually do the work alongside them. Managing campaigns, draft reports, updating CRMs, drafting emails - all within guardrails they could trust.
I built Praxis Agents because that product didn't exist. A governed AI agent platform where every action is auditable, every sensitive decision needs human approval, and your data stays private. Built for how teams of the future will operate.
Based in the UK, available globally
Praxis Agents is a product of Gregory Asquith Limited, registered in England and Wales. We're based in the UK, but can work with teams anywhere in the world.
Bring your context. We'll build the rest.
Most conversations start with a Q&A. Walk us through where the manual hours go and what ‘good’ looks like - we'll show you what that looks like in an agent-led world.UK-based founderBuilt on real workflowsBring your own AI