Description

Position: LLM Application Engineer

📍 Location: Coimbatore ( perferring Coimbatore Candidates)

💼 Experience: 5+ Years

⏳ Notice Period: Immediate to 15 Days

🏢 Work Mode: Work From Office

🧠 About the Role:

We are hiring an experienced LLM Application Engineer to work on the implementation and integration of Large Language Models (LLMs) into backend systems. This role focuses strictly on LLM application development and integration and does not involve model training or AI research.

🔑 Key Responsibilities:

📌 Implement system and user prompt templates provided by senior leadership

📌 Integrate OpenAI / Azure OpenAI APIs into backend workflows

📌 Build APIs using FastAPI or similar frameworks

📌 Handle context injection, retries, rate limits, and error handling for LLM calls

📌 Integrate LLM responses into backend workflows

📌 Apply basic guardrails such as refusals, safe responses, and fallback handling

📌 Log and monitor LLM interactions for quality and debugging

📌 Work with REST APIs and JSON-based services

📌 Manage token usage, latency, and cost control

🛠 Mandatory Skills:

🎈 Strong Python programming skills

🎈 Experience integrating OpenAI or Azure OpenAI APIs

🎈 Hands-on experience with system and user prompt implementation

🎈 Experience building APIs using FastAPI or similar frameworks

🎈 Familiarity with PostgreSQL or similar databases

🎈 Experience with Git and standard CI/CD workflows

🎯 Ideal Candidate Profile:

♦ Strong Python programming skills

♦ Experience integrating OpenAI / Azure OpenAI APIs

♦ Hands-on experience with system and user prompt implementation

♦ Experience building APIs using FastAPI or similar frameworks

♦ Familiarity with REST APIs and JSON-based services

♦ Understanding of token usage, latency, and cost control

♦ Experience with Git and CI/CD workflows

🌟 Nice to Have:

♦ Exposure to LangChain or LlamaIndex (implementation level)

♦ Experience with evaluation or testing of LLM outputs

📌 Important Note:

This role focuses on LLM application and integration only and does not involve model training or AI research.