Senior AI-ML Data Scientist

We are looking for a Senior AI/ML Data Scientist to take full ownership of model development and agent behavior, managing the process end-to-end—from problem framing and algorithm selection to fine-tuning, retrieval design, agent orchestration, evaluation, and production deployment. This role demands a hands-on expert with deep knowledge of machine learning and statistics, comfortable with designing experiments, understanding attention mechanisms, and delivering solutions within strict latency constraints. A key focus is on thoughtful agent memory design—deciding what information an agent should retain, in what format, and ensuring that retention is anchored in a governed data warehouse rather than an unstructured vector store.

This position works closely with the AI Data Engineer, who manages the data warehouse, pipelines, and indexing infrastructure. The clear division of responsibilities is that the Data Engineer owns the pipeline, schema, and data guarantees, while the Senior AI/ML Data Scientist owns the algorithm, prompt engineering, and evaluation metrics.

Key Responsibilities

- Lead the end-to-end development of AI/ML models and agent systems, including problem definition, algorithm selection, and production deployment.
- Design and implement retrieval and agent orchestration strategies with a focus on efficient memory management.
- Collaborate with data engineering teams to integrate AI models with governed data warehouses.
- Conduct rigorous evaluation and fine-tuning of models to meet performance and latency requirements.
- Develop and maintain production-grade codebases using Python and relevant ML frameworks.
- Drive experimentation, including designing statistically sound tests and analyzing results.
- Manage model lifecycle processes using tools like MLflow or Weights & Biases.
- Contribute to the architecture and deployment of AI workloads on cloud platforms, optimizing for cost and performance.
- Mentor junior team members and lead technical discovery phases and client workshops when needed.

Required Qualifications

- 5 to 10+ years of experience in ML/AI engineering, data science, or related fields, with a strong track record of deploying models at scale in production environments (including LLMs, computer vision, NLP, or multimodal systems).
- Deep understanding of machine learning algorithms and neural network theory, including optimization, regularization, attention mechanisms, tokenization, embeddings, and model internals.
- Strong foundation in statistics, experimental design, and data analysis.
- Expertise in PyTorch (primary), with additional experience in TensorFlow or JAX; familiarity with the Hugging Face ecosystem (Transformers, Datasets, TRL).
- Advanced Python programming skills suitable for production environments, along with strong SQL skills for querying dimensional data warehouses.
- Hands-on experience with agentic systems such as LangChain or LangGraph, and a well-developed approach to agent memory architecture.
- Practical knowledge of knowledge graphs, including ontology design and graph-based reasoning.
- Proven ability to deploy AI workloads on cloud platforms like AWS, Azure, or GCP, including GPU provisioning, containerization, cost optimization, and CI/CD pipelines.
- Experience with experiment tracking and model lifecycle management tools such as MLflow or Weights & Biases.

Preferred Qualifications

- Direct experience implementing cognitive architectures such as CoALA, SOAR, ACT-R, or similar frameworks in production agent systems.
- In-depth knowledge of GPU acceleration internals, including CUDA, TensorRT, and cuBLAS.
- Production experience with advanced serving and training frameworks like vLLM, NVIDIA Triton, Ray Serve/Train, DeepSpeed, or Fully Sharded Data Parallel (FSDP).
- Familiarity with AI security, governance, and compliance frameworks.
- A history of contributions to open-source AI projects or published research in relevant fields.
- Proven ability to lead technical discovery phases and conduct client-facing AI workshops.
- Working knowledge of lakehouse table formats such as Iceberg or Delta Lake to collaborate effectively with data engineering teams.

This role offers the opportunity to work at the forefront of AI research and production, shaping intelligent agent behavior with a strong emphasis on scalable, governed data architectures and cutting-edge machine learning techniques.