Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm an Applied AI & Machine Learning Intern with hands-on experience in NLP, RAG systems, and ML evaluation. I’ve built and optimized end-to-end ML pipelines across research and industry settings, with a focus on model interpretability, latency optimization, and robustness testing. I have a strong foundation in Python and production-oriented ML systems, and I thrive in collaborative environments that bridge research and real-world impact.…I'm an Applied AI & Machine Learning Intern with hands-on experience in NLP, RAG systems, and ML evaluation. I’ve built and optimized end-to-end ML pipelines across research and industry settings, with a focus on model interpretability, latency optimization, and robustness testing. I have a strong foundation in Python and production-oriented ML systems, and I thrive in collaborative environments that bridge research and real-world impact.

HENG ZHENG KAI

Data Scientist, AI Engineer, Developer, +2





I'm an Applied AI & Machine Learning Intern with hands-on experience in NLP, RAG systems, and ML evaluation. I’ve built and optimized end-to-end ML pipelines across research and industry settings, with a focus on model interpretability, latency optimization, and robustness testing. I have a strong foundation in Python and production-oriented ML systems, and I thrive in collaborative environments that bridge research and real-world impact.…I'm an Applied AI & Machine Learning Intern with hands-on experience in NLP, RAG systems, and ML evaluation. I’ve built and optimized end-to-end ML pipelines across research and industry settings, with a focus on model interpretability, latency optimization, and robustness testing. I have a strong foundation in Python and production-oriented ML systems, and I thrive in collaborative environments that bridge research and real-world impact.

Available to hire

I’m an Applied AI & Machine Learning Intern with hands-on experience in NLP, RAG systems, and ML evaluation. I’ve built and optimized end-to-end ML pipelines across research and industry settings, with a focus on model interpretability, latency optimization, and robustness testing.

I have a strong foundation in Python and production-oriented ML systems, and I thrive in collaborative environments that bridge research and real-world impact.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Work Experience

Applied AI Research Assistant at Singapore-MIT Alliance for Research and Technology

September 1, 2025 - December 1, 2025

Built scalable data pipelines for high-frequency, noisy data, ensuring robustness to missing and out-of-order events. Designed modular experimentation frameworks to support reproducible ML research and benchmarking. Collaborated with researchers to validate model behaviour across diverse data regimes.

AI Research Assistant at Singapore Management University

September 1, 2025 - December 1, 2025

Implemented and evaluated multiple changepoint detection algorithms to study model sensitivity and robustness under non-stationary data distributions. Built simulation-based evaluation frameworks to stress-test ML models under distribution shift and noisy inputs. Applied SHAP-based interpretability techniques to diagnose model behaviour and improve debuggability of ML outputs.

IBM Customer Success Manager Architect Intern at IBM

May 1, 2025 - August 1, 2025

Designed and implemented an end-to-end Retrieval-Augmented Generation (RAG) pipeline integrating vector databases with LLM-based generation for enterprise AI use cases. Benchmarked and optimized retrieval and inference latency, reducing tail latency by ~30% for real-time AI workloads. Built automated CVE ingestion and preprocessing pipelines, enabling structured knowledge retrieval for downstream LLM applications. Evaluated retrieval quality and system performance trade-offs across embedding and indexing strategies.

Product Consultant Intern at Moody's Analytics

June 1, 2024 - August 1, 2024

Built automated data validation and numerical integrity checks for large-scale financial datasets, preventing silent numerical regressions in production risk systems. Contributed to model validation pipelines, analysing coefficient stability and sensitivity to identify reliability risks.