I'm Lance. I’m an AI Product Manager at MOE (Singapore) shipping GenAI products for teaching & learning, with a focus on evaluation, safety, and adoption. Nice to e-meet you!
- 📘 AI PM Case Studies: https://github.com/Leoendithas/AI-PM-Case-Studies
- 🧪 LLM Eval Playbook: https://github.com/Leoendithas/llm_eval_playbook
- 🎙️ Lumina (voice facilitator prototype): https://github.com/adoreblvnk/lumina
At MOE, I work across product, data, design, and engineering to ship classroom-facing AI features end-to-end:
- 🚢 Shipped 4 GenAI features on a nationwide platform serving 500,000+ users
- 📈 Drove adoption to ~100k monthly active users through workshops for power users and stakeholders
- ⚙️ Operationalised prompts in production (version control, regression checks, model/provider evaluations)
- 🧪 Built LLM evaluation & safety: two-layer framework (commercial safety + education-specific harms), adversarial testing, and guardrails
- 🌏 Built across 4 languages (English, Malay, Chinese, Tamil) and 20+ subjects while maintaining accuracy
- ❓ How do we measure “learning gains” fairly when teachers start from different places?
- ❓ How do we support educators to build confidence and capability—without adding workload?
- Interests: Agentic Workflows • LLM evaluation • NLP • GenAI safety • Adoption & rollout • Learning Science
- Tools: Python • TypeScript • Streamlit • FastAPI • VS Code
- Learning: PyTorch • LangChain
- ✅ Certified Scrum Product Owner (CSPO)
- ✅ Certified ScrumMaster (CSM)
- ✅ DeepLearning.AI — Machine Learning Specialization
- ✅ GovTech — AI Champions Bootcamp
- 💬 GitHub: https://github.com/leoendithas
- 🔗 LinkedIn: https://linkedin.com/in/lance-tan