Gireesh Kumar

I design systems that make decisions, not just execute steps. Twenty-six years in software: enterprise platforms in aviation and banking — Emirates, First Abu Dhabi Bank — and now agentic AI, systems that own an outcome, learn from what worked, and escalate when they should. Before that, three years as faculty teaching programming, 1997 to 2000.

What I actually do

• Enterprise architecture — event-driven, microservices, cloud-native, serverless. Architecture Review Board, Solution Intent, ADRs, transition roadmaps. TOGAF, Agile/SAFe.
• Agentic AI systems — multi-agent design, context engineering, voice and conversational agents, multilingual deployments.
• Building — I run a small product studio in Trivandrum working on AI voice and chat agents for Indian SMBs. That product is Kooday.

Kooday and TAS Detailing Studio

I've also run a physical business, TAS Detailing Studio, since 2013 — a car detailing studio in Thiruvananthapuram. It runs on Kooday: Maya, a Kooday voice agent, calls TAS's customers after service to collect feedback in Malayalam and English. That is not a reference customer picked to make a slide look good — it is the business I know best, because a five-person company either answers its phone or it doesn't, and I have felt both sides of that.

Most AI projects disappoint for the same reason: intelligence without ownership. A tool answers a question and stops; nobody is accountable for the result. Closing that gap — building systems that own an outcome rather than narrate one — is the work I care about, whether that is an enterprise architecture review board or a voice agent calling a customer back.

Credentials

• TOGAF 9 Certified, TOGAF® Certification Program – Level 1
• AWS and Azure certified
• 26 years in software, including enterprise architecture roles at Emirates, First Abu Dhabi Bank and Encora

Based in Thiruvananthapuram, Kerala, working remotely across India and the Gulf. LinkedIn .