Co-Founder Snapdrum | CTO @KidsTilli Previous: @Google, @idcMIT, @KingsCollegeLon @NUSingapore

London
Joined May 2020
AI research is developing rapidly, & I'm committed to staying up to date with the latest research & mastering the fundamentals. I'll use this thread to tweet about my progress (inspired by @Suhail). My goal for the next 2 years is to become an expert in the field of applied AI.
1
1
21
3,399
512,000 lines of Claude Code. One missing entry in .npmignore. Full breakdown in my newsletter. Every feature flag, what KAIROS means for agent architecture, and what builders should do with this. Free to read. akshikaw.substack.com/p/the-…
1
155
Hyperagents: metacognitive self-improvement is here
1
41
190 startups. 16 worth watching. YC W26 Demo Day just wrapped and the signal is clear, the winners won't be flashy consumer apps. They'll be the infrastructure players. The ones building the rails everyone else runs on.
1
1
81
Ship 10x faster with AI agents handling every sprint stage. GStack by Gary Tan. 45.7k stars. No waitlist. 🔗 github.com/garrytan/gstack
58
Who’s hiring engineers right now? Reply with the role, location, and how to apply.
45
Build it once. Claude runs it every time.
1
40
Citations vs. Quality: Can AI judge scientific impact? 📈
23
OpenAI just built the enterprise orchestration layer
25
25 Remote AI jobs just went live at Mercor. 🚀 👇 Want the direct application links for all 25 roles? Follow me (so I can DM you) Comment "AI JOBS" > I’ll send the full list to your inbox! 📩
74
Replit Agent 4: The Evolution of Vibe Coding
1
40
Dr Aki Wijesundara retweeted
1 hour of TravisCore to build and accelerate to
90
367
79
3,820
427,746
12 Best Websites to Find Remote Jobs in 2026
26
Applied AI Engineering 2026 Edition 0 → Hired Roadmap
1
1
50
Applied AI Engineering 2026 Edition 0 → Hired Roadmap
21
We're bringing @marcklingen (CEO) and Lotte Verheyden from @Langfuse for a live session on evaluating agentic systems in production. One of Europe's best AI infrastructure companies showing you how they actually do evals.
1
1
73
If you're building agents or trying to figure out how to evaluate them properly - this is the session and it is FREE! Best for: - Anyone building agentic systems - PMs transitioning into AI Engineers - who want to see production eval patterns
1
27