AI Engineer - LLM Fine-Tuning & AI Assistant Customization

We are hiring an experienced AI Engineer / ML Specialist with deep expertise in Large Language Models (LLMs), who can fine-tune, customize, and integrate state-of-the-art models like OpenAI GPT, Claude, LLaMA, Mistral, and Gemini into real-world business applications.
The ideal candidate should have hands-on experience with foundation model customization, prompt engineering, retrieval-augmented generation (RAG), and deployment of AI assistants using public cloud AI platforms like Azure OpenAI, Amazon Bedrock, Google Vertex AI, or Anthropics Claude.Key Responsibilities:LLM Customization & Fine-TuningFine-tune popular open-source LLMs (e.g., LLaMA, Mistral, Falcon, Mixtral) using business/domain-specific data.Customize foundation models via instruction tuning, parameter-efficient fine-tuning (LoRA, QLoRA, PEFT), or prompt tuning.Evaluate and optimize the performance, factual accuracy, and tone of LLM responses.AI Assistant DevelopmentBuild and integrate AI assistants/chatbots for internal tools or customer-facing applications.Design and implement Retrieval-Augmented Generation (RAG) pipelines using tools like LangChain, LlamaIndex, Haystack, or OpenAI Assistants API.Use embedding models, vector databases (e.g., Pinecone, FAISS, Weaviate, ChromaDB), and cloud AI services.Must have experience of finetuning, and maintaining microservices or LLM driven databases.Cloud IntegrationDeploy and manage LLM-based solutions on AWS Bedrock, Azure OpenAI, Google Vertex AI, Anthropic Claude, or OpenAI API.Optimize API usage, performance, latency, and cost.Secure integrations with identity/auth systems (OAuth2, API keys) and logging/monitoring.Evaluation, Guardrails & ComplianceImplement guardrails, content moderation, and RLHF techniques to ensure safe and useful outputs.Benchmark models using human evaluation and standard metrics (e.g., BLEU, ROUGE, perplexity).Ensure compliance with privacy, IP, and data governance requirements.Collaboration & DocumentationWork closely with product, engineering, and data teams to scope and build AI-based solutions.Document custom model behaviors, API usage patterns, prompts, and datasets.Stay up-to-date with the latest LLM research and tooling advancements.Required Skills & Qualifications:Bachelors or Masters in Computer Science, AI/ML, Data Science, or related fields.3-6+ years of experience in AI/ML, with a focus on LLMs, NLP, and GenAI systems.Strong Python programming skills and experience with Hugging Face Transformers, LangChain, LlamaIndex.Hands-on with LLM APIs from OpenAI, Azure, AWS Bedrock, Google Vertex AI, Claude, Cohere, etc.Knowledge of PEFT techniques like LoRA, QLoRA, Prompt Tuning, Adapters.Familiarity with vector databases and document embedding pipelines.Experience deploying LLM-based apps using FastAPI, Flask, Docker, and cloud services.Preferred Skills:Experience with open-source LLMs: Mistral, LLaMA, GPT-J, Falcon, Vicuna, etc.Knowledge of AutoGPT, CrewAI, Agentic workflows, or multi-agent LLM orchestration.Experience with multi-turn conversation modeling, dialogue state tracking.Understanding of model quantization, distillation, or fine-tuning in low-resource environments.Familiarity with ethical AI practices, hallucination mitigation, and user alignment.Tools & Technologies:Category Tools & PlatformsLLM Frameworks Hugging Face, Transformers, PEFT, LangChain, LlamaIndex, HaystackLLMs & APIs OpenAI (GPT-4, GPT-3.5), Claude, Mistral, LLaMA, Cohere, Gemini, Azure OpenAIVector Databases FAISS, Pinecone, Weaviate, ChromaDBServing & DevOps Docker, FastAPI, Flask, GitHub Actions, KubernetesDeployment Platforms AWS Bedrock, Azure ML, GCP Vertex AI, Lambda, StreamlitMonitoring Prometheus, MLflow, Langfuse, Weights & Biases