AI is moving from experiments into the core of how businesses run. The people who matter most in that shift are the ones who can turn technical complexity into something teams can actually use.
Imran Khan is one of them. He is a TOGAF-certified enterprise architect, a Director at Standard Chartered, and an AI technology consultant with more than 19 years in technology and e-commerce. Across his work, writing and teaching, he has one consistent message: AI only creates value when it is built properly and put in the hands of the people who need it.
Building AI that survives contact with production
Imran's approach to AI is shaped by years of designing large systems. His recurring point is that taking AI from proof of concept to production takes architectural discipline, not clever prompts.
Systems, not single prompts. A single call to a large language model rarely solves a real business problem. Imran argues for compound AI systems that combine several pieces:
- Retrieval-augmented generation (RAG), so models answer from the company's own data
- Small language models (SLMs) where low latency matters
- Agent-style validation loops that check outputs before they reach users
AI raises the bar for junior engineers, it doesn't replace them. Imran's view is that the next generation of developers will do well by focusing on three skills:
- Breaking systems down into components
- Verifying code rigorously
- Translating business needs into intelligent workflows
Teaching AI to public-sector professionals from 17 countries
Imran's work goes beyond the enterprise. He was a technical trainer for the Annual International Training Course (AITC), organized by the Thailand International Cooperation Agency (TICA) with King Mongkut's University of Technology North Bangkok (KMUTNB).
Alongside fellow trainer Baran Khan, he taught public-sector professionals from 17 countries how to apply AI without writing code. The sessions covered healthcare, education, agriculture and sustainable development.
Participants worked through a five-step framework:
- Prompt: describe the problem to an AI tool
- Idea: shape it into a concrete solution
- Build: create a working version with no-code tools
- Harden: test it and make it reliable
- Pitch: present it to decision-makers
The goal was for every participant to go home with a working tool for a real challenge in their own country, not just a certificate.
Making new technology accessible
Imran writes regularly about where technology is heading, including MACH architecture, e-commerce transformation and immersive technologies. Much of what he shares is hands-on, such as a step-by-step guide to running automation workflows locally with n8n.
The common thread is accessibility. He wants professionals and enthusiasts to be able to try these tools themselves, not just read about them.
Connect with Imran
Imran shares his thinking on enterprise architecture, production AI and digital transformation on LinkedIn. Follow him there for his latest posts and guides.
