.png){kind=logo}
Yupeng Tang
Junior Machine Learning Engineer specializing in LLM systems and GPU inference
Connect with Yupeng
Yupeng already has a relationship with Reval, so a warm intro from us gets a much better response than cold outreach.
Recommended
Already have an account?
About
LLM/agent engineer who shipped a production RAG-based recommendation + explanation system that replaced a traditional recommender stack, delivering ~20% CTR lift (and +8% after a reliability iteration) with strong cold-start performance. Demonstrates strong production rigor: schema-constrained generation, typed tool calling, explicit state/orchestration, deep monitoring/feedback loops, and safe integration with messy ERP inventory/order data using normalization, idempotency, and conflict-resolution guardrails.
Experience
Education
Key Strengths
- Shipped production LLM/RAG recommender with strict grounding, reranking, and robust fallbacks
- Improved CTR ~20% vs prior recommender; strong cold-start performance
- Designed monitoring split by retrieval vs generation quality with end-to-end logging and dashboards
- Built deterministic multi-step agent pipelines using typed tools, structured schemas, and explicit state tracking
- Material reliability iteration: added reranking + minimum relevance threshold + fallback, improving CTR ~8% and reducing hallucinations
- Safely integrated agents with messy multi-ERP data via normalization, validation, idempotent writes, and conflict resolution
Browse Similar
Discover more candidates like Yupeng
Search across thousands of pre-screened, high-quality, high-intent candidates on Reval.
Search Talent