Pytorch, Python, PEFT, data pipelines, behaviour cloning, batching, evaluation harnesses, LLM systems, inference optimisation, Distillation, supervised fine-tuning, LoRA, on-device deployment, KV-cache, RL, quantization

Flask, Python, Hl7, FastAPI, embeddings, vector search, ICD-10, FHIR R4, self-hosted edge inference, quantisation, LOINC, SNOMED CT, GGUF, evaluation systems, retrieval, PDPA-style governance, LLM safety, fine-tuning, LLMs, health data standards, mobile ML, llama.cpp, vLLM, voice conversational AI pipelines, LoRA, RAG, QLoRA

Apis, Cloud deployment, Databases, Cloud Infrastructure, LLMs, Security fundamentals, Evaluation workflows, Human-in-the-loop products, Generative AI APIs, Quality metrics, Backend and systems architecture, Stateful services, Async jobs, Queues, Observability, Diffusion video models, AI orchestration