Use Case Definition, Feasibility and Implementation

• Work with clients to identify, shape and prioritize AI and agentic use cases, assessing technical feasibility, data readiness, effort, cost, and expected business value.
• Translate prioritized use cases into solution designs, reference architectures, backlogs and implementation roadmaps that delivery teams can execute against.
• Build rapid, credible prototypes and demonstrators for executive labs, value sprints and proof-of-value engagements, typically within 2 to 6 week timeboxes.
• Support the industrialization of validated prototypes into scaled delivery, including handover, documentation, and engineering enablement.
• Contribute reusable accelerators, patterns and reference implementations back into the practice asset library.

Evaluation, Guardrails and Responsible AI

• Design and implement evaluation harnesses measuring relevance, grounded-ness, factuality, robustness, safety and task completion (e.g. Ragas, LangSmith, DeepEval, custom harnesses).
• Implement guardrails, prompt-injection defenses, adversarial and bias testing, content filtering and policy enforcement across agentic workflows.
• Instrument systems for observability, cost and token analytics, latency profiling, tracing and drift detection; operate within defined budget guardrails.
• Apply responsible AI practice aligned to the EU AI Act, NIST AI RMF and ISO/IEC 42001, working with our AI governance and risk colleagues.

Client Engagement and Practice Contribution

• Act as the credible hands-on technical voice in client workshops, architecture design sessions, technology evaluations and platform selections.
• Produce high-quality technical documentation, architectural blueprints, demo scripts and enablement material for client and internal teams.
• Support pre-sales and pursuits activity through solution shaping, effort estimation, technical narrative and live demonstrations.
• Coach and mentor junior engineers and consultants; contribute to practice capability building, internal skilling and knowledge assets.

Required Skills and Experience

Agentic AI

• Demonstrable, recent hands-on delivery using at least two agentic frameworks - LangGraph, AutoGen, Semantic Kernel, CrewAI, OpenAI Agents SDK, Google ADK, MS Copilot studio /Power platform including multi-agent orchestration, tool use and MCP.

Generative AI

• Deep practical experience with LLM APIs with proficiency in one or more of Azure OpenAI, OpenAI, Anthropic Claude, Google Gemini, prompt and context engineering, function calling, structured outputs, fine-tuning and model selection trade-offs.

Programming

• Expert-level Python. Strong grounding in software engineering fundamentals - clean code, testing, version control, code review, packaging and deployment. Working proficiency in at least one of TypeScript/JavaScript, Java or C#.

RAG and Data

• Production experience with vector and hybrid search in one or more of - Azure AI Search, Pinecone, Chroma, Elasticsearch — plus embedding, chunking and knowledge-graph approaches (Neo4j).

Cloud and MLOps

• Hands-on delivery on Azure, AWS or GCP AI stacks; Docker, Kubernetes, Terraform, GitHub Actions; MLflow or equivalent for experiment tracking, versioning and serving.

Consulting Skills

• Ability to shape use cases, run technical workshops, articulate trade-offs to non-technical stakeholders and work fluently in client-facing, multi-disciplinary teams.

Education

Bachelor's or Master’s degree in computer science, Engineering, Mathematics, Statistics or a related discipline.

Certifications

Candidates are expected to hold current, verifiable industry certifications. At least one cloud AI certification is mandatory; further certifications are strongly valued.

Mandatory (one or more or Equivalent)

• Microsoft Certified: Azure AI Engineer Associate (AI-102);
• AWS Certified Machine Learning - Specialty or AWS Certified AI Practitioner;
• Google Cloud Professional Machine Learning Engineer;