Senior AI & Data Science Lead — HR & Payroll

About BIPO Established in 2010 and headquartered in Singapore, BIPO powers payroll and HR for businesses across170+ countries. With 50+ offices worldwide, we combine global compliance, deep local expertise, and scalable cloud technology to manage the complete employee lifecycle — from onboarding and benefits to payroll and workforce analytics. We're at an exciting inflection point. AI is reshaping what modern HR looks like, and we're building the infrastructure and products that will define the next generation of workforce management. The Product Innovation team is where that happens. About the Role We're hiring a Senior AI & Data Science Lead to build BIPO's AI product suite from the model layer up. This is a hands-on, high-ownership role on the Product Innovation team — you'll spend most of your time in the code: designing, building, evaluating, and shipping the AI systems that reshape payroll, HR management, and workforce analytics across our platform. This is roughly a 70/30 role. The majority of your week is technical: prototyping with LLMs, building evaluation and anomaly-detection pipelines, writing production-quality code, and partnering deeply with Engineering and Data Science on architecture. The remaining share is product leadership — setting technical direction for the AI roadmap, making sharp build-vs-defer calls, and keeping senior leadership aligned on where the AI capability is going. You'll be the most senior technical voice on AI in the room. What You'll Do Hands-On AI Engineering (~70%) • Design, build, and ship LLM-powered features across the payroll assistant, conversational HR surfaces, and anomaly-detection systems — writing production-quality code, not just specs. • Own the full AI engineering loop: prompt design, retrieval and context pipelines, fine-tuning where it earns its keep, and the deterministic guardrails that keep payroll correct. • Build and maintain rigorous evaluation harnesses — hallucination scoring, regression suites, and accuracy benchmarks — so model changes ship on evidence, not vibes. • Develop anomaly and error-detection systems over payroll and HR data, combining statistical/ML techniques with LLM-driven explainability to surface root causes. • Partner daily with Engineering and Data Science on architecture, data flows, and deployment — from prototype through production hardening, monitoring, and iteration. • Stay ands-on with the fast-moving model and tooling landscape (new models, agentic frameworks, serving stacks) and turn what's newly possible into working systems. Product & Technical Leadership (~30%) • Set the technical direction for the AI roadmap — strategic bets on where AI creates real leverage across payroll, HR, and workforce analytics — and defend the sequencing to senior leadership. • Translate ground-level understanding of payroll runs and HR workflows into a prioritized set of AI opportunities, then turn the highest-value ones into shipped capability. • Make sharp build-vs-buy-vs-defer and trade-off decisions, balancing ambition against enterprise reality: correctness, latency, cost, and compliance. • Align Engineering, Data Science, Operations, and GTM stakeholders on milestones, risks, and dependencies, communicating with high clarity and minimal noise. • Mentor engineers and data scientists on AI best practices, raising the bar for how the team builds and evaluates intelligent systems. What We're Looking For • 5+years building production ML/AI systems as a hands-on engineer or applied scientist, with a track record of shipping complex, high-impact systems end-to-end. • Strong, current coding ability (Python and the modern AI stack) — you're comfortable owning code from prototype to production and would fail a role that never touches an editor. • Deep, hands-on experience with LLM-powered or AI-native systems: prompting, RAG, evaluation, fine-tuning (LoRA/PEFT), model limitations, and designing reliable behavior around them. • Experience building evaluation, benchmarking, or anomaly-detection pipelines — you measure model quality rigorously rather than trusting outputs at face value. • Adata-informed mindset: comfortable using metrics, logs, and production signals to prioritize, validate, and course-correct. • Enough product and strategic judgment to own technical direction and defend trade-offs to senior leadership — you can lead the AI agenda, not just execute a backlog. • Comfort operating across APAC markets, enterprise stakeholders, and cross-f…