Director Agentic AI Data Science

Role: Director Agentic AI Data Science

Location: Houston, TX. Hybrid, 2-3 days onsite per week

Required

15+ years of experience across software engineering, data engineering, data science, or AI/analytics, with at least 5 7 years leading AI/ML transformation initiatives.

Proven track record leading large-scale AI or digital transformation programs at consulting firms (Deloitte, PwC, Accenture, Cognizant) or equivalent director/senior manager roles in financial services technology.

Demonstrated expertise working in a forward-deployed or embedded model, owning end-to-end solution delivery from architecture through production launch and ongoing optimization.

Hands-on technical expertise in modern AI stacks: LLMs, RAG, vector databases, cloud platforms, and ML engineering practices.

Prior experience in financial services (investment banking, capital markets, wealth management, payments, insurance) or other regulated domains (healthcare, government).

Strong communication skills: ability to translate technical AI concepts for C-suite audiences and facilitate workshops with business and risk stakeholders.

Job description:

Forward Deployed Context Engineering Lead - Director

The Forward Deployed Context Engineering Lead (Director) is responsible for designing, deploying, and governing end-to-end context and retrieval architectures that power generative AI and agentic AI solutions for strategic financial services clients. Embedded at client sites using a embedded forward deployment model, this leader translates business imperatives into production-grade AI deployments, context pipelines, tool orchestration frameworks, and evaluation systems that ensure AI workflows are compliant, reliable, and measurable against regulatory and business standards.

In highly regulated financial services environments, this role bridges AI innovation with governance frameworks including BCBS 239 (risk data aggregation), NIST AI RMF, ISO 42001, and emerging AI regulations. The holder serves as both a trusted technical advisor and strategic business partner, shaping how AI augments operations.

Key Responsibilities

Strategic Partnership & Client Leadership

Partner with CFO, CIO, COO, Chief Risk Officer, and Chief Compliance Officer to define multi-year AI transformation roadmaps, with emphasis on use case prioritization, governance alignment, and risk-adjusted ROI.

Work "forward deployed" at client sites, embedding within business units and technology teams to own end-to-end solution delivery from discovery through production and value realization.

Lead discovery workshops to uncover high-value AI opportunities, frame problem statements, and shape outcome-driven implementations in core business processes (credit decisioning, market risk, AML/CFT, wealth advisory).

End-to-End Context Architecture

Own the complete context fabric that feeds LLMs and agents: data products, RAG pipelines, vector databases, knowledge graphs, semantic layers, tool schemas, memory stores, and orchestration patterns.

Design and oversee implementation of enterprise-scale retrieval systems that integrate multiple data sources (core banking, risk repositories, regulatory data, market data, customer data warehouses) with sub-second latency and high recall.

Architect tool landscapes for agents, defining function schemas, validation rules, pre/post-execution guardrails, and escalation patterns so agents safely interact with core systems (core banking APIs, trading platforms, CRM, regulatory reporting engines).

Establish context quality and freshness standards aligned to use case sensitivity: real-time for trading contexts, hourly for compliance contexts, daily for advisory contexts.

AI & Data Governance & BCBS 239 Alignment

Embed data lineage, quality controls, and metadata management into context pipelines to satisfy BCBS 239 principles (completeness, accuracy, timeliness, clarity, granularity) and emerging AI data governance expectations.

Work with Chief Data Officer and data governance teams to ensure context data products meet regulatory lineage requirements, audit trails, and change management protocols.

Design data product contracts that codify context completeness, freshness, and accuracy SLAs and make them machine-readable for automated quality gates.

AI Evaluation & Safety Strategy

Define and operationalize comprehensive evaluation strategy covering accuracy, consistency, hallucination detection, bias, fairness, latency, cost, and regulatory compliance by use case.

Establish baseline and continuous metrics for both offline benchmarking (held-out test sets, red teaming) and online monitoring (production feedback, human review, alert escalations).

Partner with Evaluation Engineering and Risk teams to implement automated quality gates in CI/CD pipelines, blocking unsafe or regressing models/prompts/context changes from deployment.

Lead design and execution of red team exercises for high-risk use cases (credit decisioning, investment advice, transaction monitoring), including jailbreak detection, prompt injection, data leakage, and discriminatory output testing.

Responsible AI & Governance Integration

Translate NIST AI RMF, ISO 42001, EU AI Act, and internal governance policies into architecture requirements: explainability, auditability, bias monitoring, human oversight, incident response.

Collaborate with Chief Risk Officer, compliance, legal, and audit to define AI governance controls: model risk management, data governance, algorithm risk registers, impact assessments, and escalation workflows.

Establish observability and monitoring frameworks that track AI system health (quality metrics, safety indicators, regulatory drift) and provide dashboards and alerts for business and risk stakeholders.

Cloud & Platform Strategy

Own technical platform strategy for AI solutions, including cloud selection (AWS, Azure, Google Cloud Platform), data platform selection (Snowflake, Databricks, Palantir Foundry, Microsoft Fabric), and integration architecture.

Architect secure, scalable, multi-tenant AI infrastructure that meets financial services standards for security, auditability, disaster recovery, and regulatory reporting.

Design and implement MLOps, LLMOps, and model governance workflows to ensure reproducibility, auditability, and rapid, safe iteration on AI solutions in production.

Business Development & Executive Engagement

Support go-to-market strategy for AI transformation engagements: solutioning, proposals, statement of work (SOW) development, and executive presentations to C-suite.

Build thought leadership content (whitepapers, case studies, reference architectures) on context engineering, AI governance, and financial services AI transformation.

Mentor and grow a practice or delivery team of context engineers, evaluation engineers, platform engineers, and solution architects to scale repeatable AI capability.

Technical Competencies Context Engineering

LLM & Agentic Systems

Deep, hands-on understanding of LLM architecture, capabilities, limitations, and fine-tuning approaches; experience with GPT-4, Claude, LLaMA, and domain-specialized models.

Advanced expertise in multi-agent orchestration patterns: hierarchical agents, collaborative agents, tool-using agents, memory strategies, and long-horizon planning for financial workflows.

Proficiency with prompt engineering, in-context learning, chain-of-thought, and few-shot prompting for complex financial reasoning tasks (credit analysis, risk assessment, advisor-style interactions).

Context & Retrieval Architecture

- Production experience designing and implementing retrieval-augmented generation (RAG) systems, including chunking strategies, embedding models, vector databases (Pinecone, Weaviate, Milvus), and hybrid search (semantic + lexical).

- Data engineering expertise: ETL/ELT pipelines, streaming architectures (Apache Kafka, Spark Structured Streaming), data quality frameworks, and metadata management catalogs.

- Knowledge graph and semantic layer experience: designing ontologies, entity resolution, relationship extraction, and knowledge graph querying for financial contexts (counterparties, instruments, risk factors, regulatory entities).

- Experience with feature stores and data products as foundational context infrastructure; ability to define and operationalize data product contracts.

AI Governance, Safety & Evaluation

Familiarity with AI safety and evaluation techniques: benchmark design, task-specific metrics, human-in-the-loop review, red teaming, jailbreak testing, and bias/fairness audits.

Working knowledge of NIST AI Risk Management Framework, ISO 42001, BCBS 239 data governance principles, and emerging financial AI regulations (EU AI Act, Treasury AI guidance).

Experience with model risk management (MRM) frameworks, model cards, impact assessments, and governance workflows for high-risk AI systems in regulated environments.

Cloud & Data Platforms

Production architecture and deployment experience on major clouds (AWS Bedrock, SageMaker, EC2; Azure Copilot, OpenAI Service, App Service; Google Cloud Platform Vertex AI).

Proficiency with enterprise data platforms: Snowflake (architecture, Cortex AI), Databricks (LLMs, fine-tuning), Palantir Foundry (ontologies, apps), Microsoft Fabric (data engineering, AI services).

Security, compliance, and observability: encryption at rest/in transit, IAM, audit logging, HIPAA/SOX/GDPR compliance controls, and monitoring/alerting.

Software Engineering & MLOps

Strong Python and at least one of TypeScript, Java, or Scala; experience building production ML/AI systems, not just prototypes.

Hands-on MLOps, LLMOps, and CI/CD: infrastructure-as-code (Terraform, CloudFormation), containerization (Docker, Kubernetes), experimentation platforms, and model deployment pipelines.

API design and integration: designing REST/GraphQL APIs, integrating AI systems into microservices architectures, event-driven systems, and enterprise applications (core banking, trading systems).