🎁 Before you apply, rehearse this interview. Create your free WorkMundi account and get an Interview Training on HelpsYouSpeak — no cost, no card. I want my training →
About the Role: We are looking for a hands-on AI/ML Developer with experience in Large Language Models (LLMs), Prompt Engineering, and AI model integration. The ideal candidate should have practical experience working with AI models, fine-tuning them, optimizing prompts, and integrating them into real-world applications. This role is perfect for someone who has already worked on AI-driven applications and wants to expand their expertise by researching and implementing new AI advancements. You will have the opportunity to experiment with different LLM architectures, improve AI model efficiency, and contribute to AI-driven solutions. Key Responsibilities: LLM Development & Implementation: Work hands-on with LLMs like GPT, LLaMA, Mistral, Claude, and Gemini. Fine-tune models using Hugging Face Transformers, PyTorch, or TensorFlow. Train, optimize, and deploy LLMs for tasks like text generation, summarization, and chatbots. Prompt Engineering & Optimization: Design, test, and optimize prompts for various AI tasks. Apply Few-shot, Zero-shot, Chain of Thought (CoT), and ReAct prompting. Improve AI model accuracy by iterating and refining prompt strategies. AI Model Integration & Deployment: Develop Python-based applications that interact with LLM APIs. Build and deploy AI models using FastAPI or Flask. Work with vector databases (FAISS, ChromaDB, Pinecone) for efficient retrieval. Deploy AI models on cloud platforms (AWS, Azure, GCP) using Docker and Kubernetes. Data Processing & NLP: Preprocess and clean large-scale text datasets. Work with text embeddings, named entity recognition (NER), and knowledge retrieval. Implement vector search techniques for AI-enhanced applications. AI Research & Experimentation: Stay up to date with the latest LLM advancements and AI research papers. Implement recent AI techniques into existing workflows. Optimize models using quantization, vLLM, and low-rank adaptation (LoRA). Required Skills & Experience: Must-Have Hands-on Experience: Python programming with AI/ML frameworks (Hugging Face, PyTorch, TensorFlow). Hands-on experience working with LLMs and fine-tuning. Experience in prompt engineering and optimizing AI model outputs. Building APIs with FastAPI or Flask for AI model integration. Familiarity with vector databases and embedding models. Nice to Have (or Learn on the Job): Experience with LangChain, LlamaIndex, or Retrieval-Augmented Generation (RAG). Knowledge of quantization techniques (LoRA, GPTQ, vLLM, ONNX) for efficient model deployment. Experience working with knowledge graphs and reasoning-based AI. Background in MLOps for tracking and managing AI models. Location : - Mumbai, Delhi / NCR, Bengaluru , Kolkata, Chennai, Hyderabad, Ahmedabad, Pune, Remote .
Here's where they are and how to stand out in your next interview.
24.795Jobs
9.577IN
74%EN
That number includes positions in India (9,577), Poland (3,429), the United States (1,753), Singapore (787), and Brazil (779). If you're looking to move or work remotely, check which countries match your goals — then filter WorkMundi by location to see what's actually hiring near you.
Most developer ads you'll see are written in English: 18,397 of the 24,795. That doesn't tell you whether the job itself requires you to speak English or work in English daily, so read each posting carefully. Don't assume the language of the ad matches the language of the team.
The employers posting the most developer roles are Link Group (369 jobs), Upvanta (257), jobgether (226), and OfferZen (212). If you're applying to any of these, research their hiring patterns. They move fast and post often, which means they're either scaling hard or replacing people who didn't fit.
In your interview, expect this: 'Walk me through the last time you had to debug something that took you more than an hour. What was it, what did you try first, and what would you do differently?' Hiring managers ask this to see if you think systematically or just try random fixes. Have a real example ready with specifics.