Must Have Skills
Skill 1 – Design end-to-end ML/LLM and agentic solutions, from problem framing and data strategy through to deployment and monitoring
Skill 2 – Architect agentic systems: multi-step, tool-using, and multi-agent workflows: including orchestration, tool/function integration, memory, and guardrails
Skill 3 – Work with foundation models including Claude and other LLMs: prompting, fine-tuning, evaluation, and integration
Key Responsibilities
Solution Design & Technical Leadership:
• Design end-to-end ML/LLM and agentic solutions, from problem framing and data strategy through to deployment and monitoring.
• Architect agentic systems: multi-step, tool-using, and multi-agent workflows: including orchestration, tool/function integration, memory, and guardrails.
• Own the technical architecture for solutions: model selection, fine-tuning approach, agent/orchestration design, serving strategy, API design, and infrastructure footprint.
• Lead and mentor a team of data scientists and developers, break complex, ambiguous customer requirements into structured project plans with clear milestones and deliverables. AI/ML
Required Qualifications:
• Bachelor's & Master's degree in Computer Science, Data Science, Machine Learning, or a related field (or equivalent practical experience).
• Experience in agentic frameworks and protocols (e.g., LangGraph, LlamaIndex, AutoGen, CrewAI, MCP) and with RAG and tool-use patterns.
• Familiarity with MLOps tooling (experiment tracking, CI/CD for ML, model registries, monitoring).
• Experience with distributed training and inference optimization (quantization, batching, GPU utilization).
• Exposure to containerization and orchestration (Docker, Kubernetes).
Technical Frameworks & Toolkit: Deep Learning frameworks: PyTorch, TensorFlow, Keras, JAX; PyTorch Lightning. GenAI & fine-tuning frameworks: Hugging Face Transformers, PEFT (LoRA/QLoRA), TRL, Accelerate, DeepSpeed, bitsandbytes, Axolotl, Unsloth; vLLM / TGI / Ollama for serving; LangChain, LlamaIndex for RAG and orchestration. Agentic AI frameworks & protocols: Claude Agent SDK, Anthropic / OpenAI SDKs, LangGraph, AutoGen, CrewAI, Semantic Kernel, and the Model Context Protocol (MCP); tool/function calling and multi-agent patterns. Computer vision: OpenCV, Detectron2, Segment Anything (SAM); image/video pipelines. Forecasting & optimization: stats models, Prophet, GluonTS, Darts, scikit-learn; optimization/solver tooling (e.g., OR-Tools, SciPy, PuLP, Gurobi/CVXPY). MLOps & infra