Senior ML Developer at Coveo Solution Inc.

Ranking and recommendation systems for e-commerce using ML, LLM, and GenAI; building ML infrastructure and optimizing deployment; PEFT finetuning, prompt engineering, and prompt-tuning of SOTA models; deploying LLMs through Bedrock and self-hosted services; migrating production services to microservice architecture with Python FastAPI; data integration with Snowflake; leveraging AWS SageMaker; enabling structured LLM outputs with function calling and constraint sampling; building internal workflow agents with MCP, tools, and LLM function calls; implementing guardrails for responsible GenAI.

Sr. ML Engineer at MeetKai

Built enterprise virtual assistant system with SOTA NLP and conversational AI technologies: intent classification and entity recognition, multi-turn conversation, context and memory management, and more. Built PoC virtual assistant system with RASA. Built cloud backend for ConvAI with semantic search (dense search with BM25, semantic search with vector databases such as ElasticSearch, Pinecone, Weaviate) and RAG pipelines (canonical RAG, GraphRAG). Fine-tuned BERT models for RAG pipelines and adapted various LLMs with transfer learning and fine-tuning. Developed Python-based cloud RESTful microservices with FastAPI; integrated Redis and PostgreSQL via containerization. ML DevOps to build, release, deploy, and maintain ML cloud products with Docker, Kubernetes, GitHub CI/CD, and GCP/AWS clouds. Fine-tuned function-calling specialized LLM. Implemented guardrails for responsible GenAI including profanity detection and alignment.