Context
AI work that goes beyond demos: systems running in production, handling real queries, embedded in daily operations. Built inside 
Iterativeiterative.vcEarly-stage VC fund investing up to US$500K in startups twice a year, working closely with founders for 3 months — 70+ portfolio companies across Southeast Asia and South Asia.
Iterative, a SEA-focused VC fund, and shaped by the only test that matters — whether a real team keeps using it after the novelty wears off.
What I Did
- AI-Powered Deal Tools: Built a custom MCP (Model Context Protocol) server giving the team natural-language access to Airtable, Notion, Slack, and internal data. Evolved through multiple major versions; now handles hundreds of queries per month.
- Voice AI Pipeline: Designed a voice cloning pipeline using pyannote.audio and the Grain REST API to create
ElevenLabselevenlabs.ioAI voice generator and voice agents platform — 5,000+ voices in 70+ languages with secure APIs and SDKs.ElevenLabs Professional Voice Clones for 18 team members. - Agent Architecture Evaluation: Evaluated multiple architectures (ElevenLabs+Vapi, Voiceflow, ElevenAgents) for automating founder outreach — and made the build-vs-buy call with real cost and latency data, not vendor decks.
Stack
MCPmodelcontextprotocol.ioModel Context Protocol — the open standard for connecting AI models to external tools, data sources, and systems. The backbone of production agent integrations.MCPPythonpython.orgThe lingua franca of data and AI engineering — used here for ML pipelines, voice processing, and backend services.Pythonpyannote.audiogithub.comOpen-source speaker diarization toolkit built on PyTorch — identifies who spoke when in audio recordings, the first stage of any voice cloning pipeline.pyannote.audioElevenLabselevenlabs.ioAI voice platform — text-to-speech, voice cloning, and conversational agents with 5,000+ voices across 70+ languages.ElevenLabsVapivapi.aiDeveloper platform for building, testing, and deploying real-time voice AI agents over phone and web.Vapin8nn8n.ioSource-available workflow automation — node-based, self-hostable, and the workhorse behind most of the operational automations I ship.n8n
Outcome
- Hundreds of AI queries processed monthly via an MCP server in production
- 18 professional voice clones created and deployed for the team
- AI tooling embedded in the fund's daily operations — not shelfware