I've spent 10+ years building software, and these days I'm fully focused on one thing: agentic AI that actually does the work instead of just chatting about it.
I build intelligent agents with OpenAI and Anthropic Claude that take real actions. Think agents that navigate an internal dashboard reliably, tools that turn a multi-hour research task into a few minutes, and assistants that plug into the systems your team already uses. I work heavily with Claude Computer Use and tool calling, so the agents I build can read, decide, and follow through, not just generate text.
On the automation side, I build production workflows with n8n, including self-hosted deployments on cloud infrastructure. At Oracle I deployed n8n on Oracle Cloud, mapped out where automation would save the most time, and built a large library of workflows that connect internal systems end to end.
My background spans Oracle, Honey Health, SingHealth, and Carousell, so I move comfortably between fast product work and reliable internal systems.
What you can count on:
✅ Agents and automations built for reliability and safety, not just demos
✅ Expert in Claude Computer Use, tool calling, and n8n
✅ Clear communication, detailed documentation, and a clean handoff
If you've got a bottleneck that a smart agent or a solid automation could fix, let's talk about it this week.