Lead AI/ML and MLOps Consultant

Strong grounding in the full ML lifecycle: data pipeline creation, feature engineering, model training, evaluation, deployment, and monitoring.

Production experience designing and building data pipelines that feed ML workloads (batch and streaming).

Solid hands-on understanding of model training: hyperparameter tuning, validation strategies, dealing with class imbalance, leakage, common failure modes.

Ability to select appropriate model families (classical ML, deep learning, large language models) for the problem at hand and justify the choice.

Production experience building and operating ML systems on at least one major cloud: GCP, AWS, or Azure.

Strong comfort with the data and AI services on that cloud (e.g. BigQuery / Vertex AI, Redshift / SageMaker, Synapse / Azure ML).

Cross-cloud experience and the ability to make pragmatic platform recommendations is a strong plus.

Comfortable in a consulting setting: multiple concurrent engagements, ambiguity, scoping under time pressure, and frequent client interaction.

Strong written and verbal communication, able to hold a technical conversation with a CTO and explain a model decision to a non-technical or business stakeholder in the same hour.

Prior experience supporting pre-sales activity (scoping, technical proposals) is strongly preferred.

Comfortable being on camera and in the room with prospects and partners.

Experience in regulated industries (healthcare, life sciences, financial services, public sector).

Production experience with RAG, vector search, and LLM evaluation frameworks.

Open-source contributions, public talks, or technical writing.

Prior experience working inside a cloud or data partner ecosystem (MongoDB, GCP, AWS, Azure, Databricks, Snowflake).

Practical experience with model explainability techniques: SHAP, LIME, feature attribution, partial dependence, model cards.

Familiarity with responsible AI practices: bias evaluation, fairness, calibration, uncertainty quantification, and confidence-aware UX patterns (e.g. withholding low-confidence predictions).

Awareness of what it takes to make a model trustworthy in regulated or high-stakes domains.

Hands-on experience designing and shipping agentic AI solutions in production or production-adjacent settings.

Strong understanding of common agent design patterns, ReAct, plan-and-execute, tool use, reflection, multi-agent orchestration, human-in-the-loop.

Working experience with one or more agent frameworks (e.g. LangChain / LangGraph, LlamaIndex, CrewAI, etc.) and vector databases.

Sound judgement on when an agent is the right tool, and when a simpler approach is.

Strong working knowledge of modern data platforms, relational, NoSQL, warehouse, and lakehouse.