Open to opportunities

Yash Halwai

@yashhalwai

Message

AI Engineer building GenAI systems: LLM orchestration, RAG, and multi-agent automation.

India

Message

What I'm looking for

I’m looking for a team where I can build and scale production LLM/RAG + multi-agent systems, with strong emphasis on measurable accuracy, observability, and cost/performance optimization across enterprise deployments.

I’m an AI Engineer focused on shipping production GenAI systems—especially LLM orchestration, RAG, agentic workflows, and AI microservices. I’ve contributed to 3 live products and 12+ enterprise deployments, with results like 95%+ accuracy and ~60% cost reductions.

At i2e Consulting (Life Sciences & AI), I lead backend development for an AI-powered portfolio planning portal for a global biopharmaceutical client. I support intelligent automation across PPM consulting, AI/ML systems, clinical data management, and custom enterprise analytics for decision support.

In my recent role at Benchmark IT Solutions, I architected 15+ AI validation microservices (FastAPI) and built a multi-provider LLM “factory” pattern across AWS Bedrock, Azure AI, and OpenAI. I delivered 95%+ text validation accuracy and cut manual review time from 2–4 hours to 3–5 minutes, while driving reliability with distributed task queues and end-to-end observability.

I also built agentic and enterprise-grade pipelines—like a 12-agent LangGraph workflow and Temporal distributed workflows for HITL approvals and vector-based semantic duplicate detection. My engineering approach emphasizes deterministic orchestration, measurable cost/performance wins, and traceable experimentation using tools like Langfuse and Helicone.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

I2E Consulting

May 2026 - Present (1 month)

Led backend development of an AI-powered portfolio planning portal for a global biopharmaceutical client, enabling executive-level review and decision support across enterprise pipelines. Contributed to life-sciences AI systems spanning AI/ML automation, clinical data management, and custom analytics solutions.

RAG Agentic AI fastAPI Document Intelligence Voice AI Async Await

Software Engineer - AI Products

Benchmark IT Solutions

Jun 2025 - May 2026 (11 months)

Architected AI validation and test automation systems using deterministic pipelines and multi-agent orchestration, improving accuracy and reducing manual review time. Built distributed workflows and semantic retrieval components to support high-throughput LLM inference with measurable uptime and cost reductions.

fastAPI LangGraph LiteLLM Celery Qdrant LLM Observability

Associate Software Engineer

Benchmark IT Solutions

Jun 2024 - May 2025 (11 months)

Developed serverless and async document/AI pipelines for job matching, OCR preprocessing, and structured extraction at scale. Implemented provider-agnostic LLM routing and multi-agent/RAG systems, including voice and JWT-secured document agent APIs for enterprise use.

AWS Lambda OpenSearch Bedrock Celery LLM Provider Abstraction

Software Engineer Intern

Benchmark IT Solutions

Feb 2024 - May 2024 (3 months)

Built production-minded AI assistants and chatbots, including CRM response automation and healthcare policy Q&A, achieving high actionable resolution rates and automated task completion. Containerized and delivered conversational AI solutions using LLMs, retrieval, and telephony integrations.

GPT 4o Gemini Faiss Docker REST APIs WebSockets