Data Science Manager – VoiceAI
Hands-on technical leadership role owning Trusting Social's VoiceAI platform — the real-time ASR→LLM→TTS pipeline behind AI voice agents for collections and sales calls. The manager sets the technical roadmap, personally drives model development (TTS fine-tuning, AMD, end-of-turn detection) in Vietnamese, Tagalog, and US English, and leads a team of data scientists and engineers.
- Compensation
- Not disclosed
- Location
- Ho Chi Minh City, Vietnam
- Work setup
- Not specified
- Seniority
- Lead
- Employment
- Not specified
- Experience
- 2+ years
- Category
- Data science
- Posted
Skills
Must have
- 6+ years in data science or applied ML
- 2+ years leading a team or acting as technical lead
- Deep expertise in speech or audio ML (ASR, TTS, VAD, speaker modeling) with models shipped to production
- Full-stack mindset: own the lifecycle from modelling and engineering to deployment and production monitoring
- Real-time mindset for voice: design for streaming and low latency, think in milliseconds
- Keeps up with SOTA in speech and LLM research and vendor market, distinguishing production-ready from hype
- Proven track record taking models from research to production and measuring business impact
- Strong statistics and evaluation fundamentals: experiment design, calibration, error analysis, FPR/FNR trade-offs
- Fluent in Python and SQL; comfortable reviewing production code and ML pipelines
- Ability to make and defend technical decisions with data and communicate clearly with engineers and executives
- AI-native way of working: use AI tools (coding agents, LLM assistants) daily
Nice to have
- Experience post-training or fine-tuning large speech or LLM models (SFT, preference optimization, distillation)
- Hands-on experience with real-time voice stacks (LiveKit, Pipecat) and GPU model serving
- Experience building a team from early stage or through an organizational transition
Full description
##
Location: Ho Chi Minh City, Vietnam,,
VoiceAI builds the real-time voice layer behind Trusting Social's AI agents, for both outbound and inbound calls. We own the cascade pipeline that connects the caller to the agent (ASR → LLM → TTS): real-time speech understanding (ASR), real-time speech generation (TTS), the real-time orchestration between speech and the LLM script, and the interaction behaviour on top, including turn-taking, latency, and answering machine detection. Our platform powers Kompato (AI collections for US fintech lenders) and Sophia (financial services sales in Vietnam and the Philippines), working closely with the Collin and Sophia DS and product teams, who own the LLM dialogue scripts, and the engineering team, who owns telephony and inbound routing and connects to our layer through the same interface for inbound and outbound calls.
About the role
This is a hands-on technical leadership role. You will set the technical direction for the VoiceAI platform, lead a team of data scientists and engineers, and personally drive the development of the models and evaluation systems that define our voice quality. We are looking for someone who can go deep on the technical details and also lead people, roadmap, and stakeholders.
What you'll do
- Own the technical roadmap for the VoiceAI platform: cascade pipeline orchestration (ASR → LLM → TTS), real-time speech understanding, real-time speech generation, turn-taking, latency, and answering machine detection, for outbound and inbound
- Lead architecture and build-vs-buy decisions, such as cascade vs. speech-to-speech, vendor vs. self-hosted speech models, and quality/cost/latency trade-offs
- Stay on top of the state of the art in speech understanding, speech generation, and speech-to-speech models (research, open-source releases, and commercial vendors), quickly benchmark what is new, and turn it into roadmap decisions
- Drive model development hands-on: TTS post-training and fine-tuning, AMD, end-of-turn detection, and speech model adaptation for Vietnamese, Tagalog, and US English
- Define the team's evaluation standard: offline benchmarks, TTS arenas, audio-based judges, human rating programs, and online A/B testing
- Lead, hire, and mentor data scientists and VoiceAI engineers; set clear goals, review work, and raise the technical bar
- Run delivery: planning, prioritization, OKRs, and execution against production reliability and quality targets
- Partner with the Collin and Sophia DS and product teams, engineering, and senior leadership; present evidence-first recommendations to the CTO and business stakeholders
What we're looking for
- Hands-on with a full-stack mindset: you own the whole lifecycle, from modelling and engineering to deployment and production monitoring