Kevin Hwang
Details
Skills
Applied AI, Artificial Intelligence, Machine Learning, Generative AI, Large Language Models, LLMs, Retrieval-Augmented Generation, RAG, Enterprise Search, Hybrid Retrieval, Hybrid Search, Semantic Search, Information Retrieval, Ranking Systems, Reranking, Recommendation Systems, Multi-Agent Systems, AI Agent Orchestration, Tool-Using AI Agents, Human-in-the-Loop AI, Permission-Aware Retrieval, Knowledge Graphs, Graph-Based Retrieval, Python, Java, SQL, LangGraph, MCP, LlamaIndex, OpenAI, Anthropic Claude, Llama, PyTorch, Hugging Face, Elasticsearch, BM25, Sentence Transformers, FAISS, pgvector, Neo4j, Redis, PostgreSQL, Kafka, Kafka Connect, Apache Spark, Ray, FastAPI, gRPC, AWS, GCP, Kubernetes, Amazon EKS, Docker, Helm, Terraform, OpenTelemetry, Prometheus, Grafana, Vector Search, Embeddings, Multimodal Retrieval, ML Ranking, A/B Testing, Causal Inference, Experimentation, Regression Adjustment, Variance Reduction, Cohort Analysis, Treatment-Effect Analysis, NDCG, Recall@K, Precision@K, MRR, MAP, LLM Evaluation, LLM-as-Judge, AI Observability, Hallucination Analysis, Retrieval Evaluation, ACL Enforcement, Data Pipelines, Distributed Systems, CI/CD, Automated Testing, Production Monitoring, Forward Deployed Engineering, Customer Cloud Deployment.
About
Kevin Hwang is a Senior Applied AI Engineer and Forward Deployed Engineer with extensive experience building and deploying production AI, machine learning, retrieval, and data systems across enterprise SaaS, social media, and online marketplace environments. At Glean, he architects permission-aware enterprise retrieval and production RAG platforms, including hybrid search, reranking, multi-agent reasoning, tool-using AI agents, and large-scale ingestion systems processing more than 100 million documents. His work incorporates Python, LangGraph, MCP, LlamaIndex, OpenAI and Anthropic models, Elasticsearch, FAISS, pgvector, Neo4j, Kafka, AWS, Kubernetes, and Terraform, with a strong focus on permissions, observability, evaluation, and production deployment into customer cloud environments. Previously at Meta, Kevin worked on News Feed ranking, experimentation, causal inference, personalization, Marketplace discovery, multimodal retrieval, and recommendation systems operating at massive consumer scale. His earlier software engineering experience at Google involved large-scale batch and near-real-time data pipelines, cloud infrastructure, automated testing, CI/CD, and production monitoring.