Triệu Ngọc Tâm

Product @ BIVA — Voice Bots & Agentic Systems in Production

Hanoi, Vietnam

tamtn.dynamite.work@gmail.com|dynamite-home.pages.dev|github.com/trieungoctam|linkedin.com/in/trieungoctam

Builds AI products that ship: voice bots handling real customer calls for Vietnamese transport companies, and agentic booking systems that process transactions end to end. Cares less about which model is used than about whether the product works for the person on the other end of the call. Open to conversations about AI product roles, voice AI, and turning agent demos into products with real users.

Home ↗︎Blog ↗︎GitHub ↗︎LinkedIn ↗︎

Selected Work

Voice AI in production — Designed and deployed callbots for intercity bus operators — booking, routing and Vietnamese-language customer service. Owned the full loop: conversation design, prompt architecture, evaluation frameworks, and iteration from real call transcripts.
BIVA — Voice-based booking product built on a verify-gated completion architecture: fail-closed admission control, so the bot never confirms what it cannot guarantee.
Evaluation as a product discipline — Structured audit frameworks — lint, simulate, refine, capture lessons — with prompts treated as product surfaces: versioned, tested, observable.

Experience

BIVA

Product Builder · HanoiSep 2025 – Present

Building AI products at BIVA: voice bots and agentic systems running in production.

Minh Phuc Transformation JSC (MPT)

AI Software Engineer · VietnamJul 2025 – Oct 2025

Icetea Labs

AI Software Developer · Hanoi, VietnamDec 2023 – Jul 2025

Programming PTIT Club

MemberDec 2021 – Aug 2024

Technical Skills

Top Skills: Agents, Multi-agent Systems, Agent booking

AI Product: Voice AI / callbots, Conversation design, Prompt architecture, Evaluation frameworks, AI product management

This Site Is Built With: TypeScript, Rust, React 19, TanStack Start, Hono, Cloudflare Pages & Workers, WebAssembly, Vite, Turborepo, Biome

How I Work

Start from the user problem, not the tech. A bot that resolves 80% of calls beats a clever architecture that resolves 60%.
Ship, measure, iterate. Transcript-observable metrics over dashboards nobody reads.
Bridge business and engineering: unit economics with stakeholders, token economics with engineers.