Nlp, Pytorch, Python, Prompt engineering, RAG architecture, transfer learning, production AI solutions, Hugging Face, LLM fine-tuning techniques, model deployment, PEFT adapters, LoRA, Transformers, LLM orchestration, Transformer architectures

Data Factory, Emr, Data Warehousing, Tensorflow, DataFlow, Deep Learning, Etl, Pytorch, Gcp, ELT, S3, Apache Spark, AWS, Python, Azure, Nlp, Generative AI, LLMs, HDInsight, LlamaIndex, Hugging Face, LangChain, Pinecone, ChromaDB, RAG architectures, FAISS

memory allocation policies, parallel processing architectures, profiling code, neural network architectures, NPU micro-architecture, TensorFlow Lite, ONNX Runtime, performance monitoring tools, model optimization and compression techniques, deep learning fundamentals, NPU-specific compilers, Windows Performance Toolkit, PyTorch Mobile

Cuda, Ubuntu, Python Scripting, Tensorflow, Deep Learning, Test automation, Yocto, Pytorch, Opencl, Performance profiling, Optimization, hip, Linux expertise, ROCm validation

Cuda, Python Scripting, Tensorflow, Test automation, Pytorch, Opencl, Deep Learning, Yocto, Ubuntu, Linux, hip, LLMs, ROCm

PostgreSQL, Docker, FastAPI, Python, Kubernetes, LangChain, embedding strategies, pgvector, RAG pipeline architecture, prompt lifecycle management, vector database selection, Azure OpenAI, LLM engineering, LangGraph, retrieval strategies, Langfuse, context management strategies

developing production-grade backend services using Python, observability and monitoring tools, relational and non-relational databases, infrastructure-as-code, cloud provisioning frameworks such as Terraform or CloudFormation, identity and access management technologies, container orchestration such as ECS Fargate, secrets management, pipeline automation with Azure DevOps, networking and load balancing, LLM API gateway technologies, asynchronous programming frameworks, multi-provider integration across major cloud AI platforms, containerisation with Docker