Senior Pyhton AI Developer
Your Responsibilities
- Develop, train, and optimize ML models using PyTorch, TensorFlow, and Keras.
- Build end-to-end LLM and RAG pipelines using LangChain and LangGraph.
- Work with LLM APIs (OpenAI, Anthropic Claude, Azure OpenAI) and implement prompt engineering strategies.
- Utilize Hugging Face Transformers for model fine-tuning and deployment.
- Integrate embedding models for semantic search and retrieval systems.
- Work with transformer-based architectures (BERT, GPT, LLaMA, Mistral) for production use cases.
- Implement LLM evaluation frameworks (RAGAS, LangSmith) and performance optimization.
- Implement real-time communication with FastAPI WebSockets.
- Implement pgvector for embedding storage and similarity search with efficient indexing strategies.
- Integrate vector databases (pgvector, Pinecone, Weaviate, FAISS, Milvus) for retrieval pipelines.
- Containerize AI services with Docker and deploy on Kubernetes (EKS/GKE/AKS).
- Configure AWS infrastructure (EC2, S3, RDS, SageMaker, Lambda, CloudWatch) for AI/ML workloads.
- Version ML experiments using MLflow, Weights & Biases, or Neptune.
- Deploy models using serving frameworks (TorchServe, BentoML, TensorFlow Serving).
- Implement model monitoring, drift detection, and automated retraining pipelines.
- Build CI/CD pipelines for automated testing and deployment with 80% test coverage (pytest).
- Follow security best practices for AI systems (prompt injection prevention, data privacy, API key management).
- Participate in code reviews, tech talks, and AI learning sessions.
- Follow Agile/Scrum methodologies and Git best practices.
Required Qualifications
- Bachelor's or Master's degree in Computer Science, AI/ML, or related field.
- 25 years of Python development experience (Python 3.9+) with strong AI/ML background.
- Hands-on experience with LangChain and LangGraph for building LLM-powered workflows and RAG systems.
- Deep learning experience with PyTorch or TensorFlow.
- Experience with Hugging Face Transformers and model fine-tuning.
- Proficiency with LLM APIs (OpenAI, Anthropic, Azure OpenAI) and prompt engineering.
- Strong experience with FastAPI frameworks.
- Proficiency in PostgreSQL with pgvector extension for embedding storage and similarity search.
- Experience with vector databases (pgvector, Pinecone, Weaviate, FAISS, or Milvus).
- Experience with model versioning tools (MLflow, Weights & Biases, or Neptune).
- Skilled in Git workflows, automated testing (pytest), and CI/CD practices.
- Understanding of security principles for AI systems.
- Excellent communication and analytical thinking.
Nice to Have
- Experience with multiple vector databases (Pinecone, Weaviate, FAISS, Milvus).
- Knowledge of advanced LLM fine-tuning (LoRA, QLoRA, PEFT) and RLHF.
- Experience with model serving frameworks and distributed training.
- Familiarity with workflow orchestration tools (Airflow, Prefect, Dagster).
- Knowledge of quantization and model compression techniques.
- Experience with infrastructure as code (Terraform, CloudFormation).
- Familiarity with data versioning tools (DVC) and AutoML.
- Experience with Streamlit or Gradio for ML demos.
- Background in statistics, optimization, or applied mathematics.
- Contributions to AI/ML or LangChain/LangGraph open-source projects.
