3780013-Lead Assistant Manager

EXL; Dadri; On-site; INR 1,500,000 - 2,500,000; Full time

Job summary

EXL is seeking an AI/ML Engineer to fine-tune open-source LLMs, build AI agents, and implement hybrid LLM solutions. You will develop Python APIs with FastAPI, containerize with Docker, and deploy on AWS/Azure while collaborating with senior engineers to optimize performance.

The role requires hands-on CUDA toolkits experience, familiarity with LangChain-like frameworks, and strong knowledge of vector databases and cloud deployments.

Qualifications

• Hands-on experience with LLM fine-tuning and CUDA toolkits.
• Familiarity with LangChain-like agent frameworks.
• Experience developing APIs with FastAPI and deploying via Docker.
• Proficiency in Python, cloud deployment (AWS/Azure), and vector DBs (FAISS, Pinecone).

Responsibilities

• Fine-tune and adapt open-source LLMs (e.g., LLaMA 4, Mistral and Bert) using NVIDIA GPU tools.
• Build AI agents using LangChain, LangGraph, or AutoGen with structured workflows (memory, tools, retries).
• Implement hybrid LLM solutions using OpenAI/Claude APIs and open-source models.
• Develop APIs using FastAPI and containerize apps with Docker.
• Deploy, monitor, and scale AI solutions on AWS, Azure, or similar cloud providers.
• Collaborate with senior engineers to optimize performance and reliability of deployed systems.

Skills

Python

Problem solving

Communication

Independent work

Tools

LangChain

LangGraph

AutoGen

FastAPI

Docker

CUDA

OpenAI APIs

Anthropic APIs

FAISS

Pinecone

LangServe

Semantic Kernel

GitHub Actions

Prometheus

Job description

Key Responsibilities:

Fine-tune and adapt open-source LLMs (e.g., LLaMA 4, Mistral and Bert) using NVIDIA GPU tools. Build AI agents using frameworks like LangChain, LangGraph, or AutoGen with structured workflows (memory, tools, retries, etc.). Implement hybrid LLM solutions using OpenAI/Claude APIs and open-source models. Develop APIs using FastAPI and containerize apps with Docker. Deploy, monitor, and scale AI solutions on AWS, Azure, or similar cloud providers. Collaborate with senior engineers to optimize performance and reliability of deployed systems.

Requirements:

Hands-on experience with LLM fine-tuning and NVIDIA GPU toolkits (CUDA). Familiarity with LangChain or similar agent frameworks. Experience developing APIs with FastAPI and deploying via Docker. Proficiency in using OpenAI/Anthropic APIs and building basic RAG pipelines. Solid foundation in Python, cloud deployment (AWS/Azure), and vector databases (e.g., FAISS, Pinecone). Nice to have: Exposure to tools like LangServe and Semantic Kernel Familiarity with CI/CD pipelines and monitoring tools (e.g., GitHub Actions, Prometheus). Contribution to open-source AI/ML projects.

Soft Skills:

• Strong communication skills - both verbal and written


• Excellent problem-solving and debugging skills


• Self-motivated with the ability to work independently and in a team


• Comfortable working with stakeholders across different time zones

Working Hours:

General Shift: 1:30 PM to 11:30 PM IST. Flexibility to extend hours based on critical deployments or support needs.