Principal Forward Deployed Data Scientist

Barclays•New York, NY

7d•$235,000 - $285,000

About The Position

This is a very senior, hands-on technical leadership role focused on building production AI applications for internal teams and customers in a regulated environment. The role works at the intersection of the business, users, and emerging AI technology, partnering directly with stakeholders to clarify ambiguous problems, choose suitable approaches, build solutions, deploy them, and support them as they evolve. The Principal Forward Deployed Data Scientist takes AI ideas from prototype through production, designs and operates agentic AI systems, and establishes technical standards for agentic AI delivery across the organisation. The role requires sound judgement in selecting models and architectures, delivering reliable software within governance and compliance controls, and keeping pace with emerging AI technologies. The position also carries leadership expectations, including demonstrating Barclays' LEAD behaviours and its Values and Mindset.

Requirements

Experience building AI/LLM applications in production that internal or external users rely on.
Proven ability to move ideas from proof of concept to hardened, scalable, maintainable production systems.
Proven record of delivering lasting-impact software to production in a regulated environment, working within governance, risk, compliance, and audit controls.
Hands-on experience building agentic AI systems involving reasoning, planning, tool use, actions, multi-agent and agent-to-agent coordination, MCP-based tool and data integration, and controls for trustworthy autonomous behaviour.
Knowledge of identity and access for AI agents, including authentication, delegated authority to act on behalf of users, non-human or service identities, authorisation, and auditability.
Strong software engineering fundamentals, including building reliable, observable, testable, and maintainable systems.
Excellent communication and stakeholder skills, including the ability to move between business discussions and technical design and work directly with non-technical stakeholders.
A forward-deployed mindset: comfort with ambiguity, proximity to business and users, and a bias toward shipping and iterating.

Nice To Haves

Experience with retrieval-augmented generation (RAG) to ground responses in trusted organisational knowledge.
Experience defining, measuring, and improving AI and agent behaviour through evaluations.
Experience with guardrails and safety, including input/output filtering, hallucination mitigation, controls over autonomous actions, and responsible-AI practices.
Experience with prompt and context design, memory, and techniques for shaping reliable model and agent behaviour.
Range and judgement in selecting models, agent patterns, and architectures to fit different problems.
Experience leading architecture across teams and setting technical direction for other engineers.
Strong interest in new and emerging technology, with the ability to distinguish production-ready technology from hype.
Experience building customer-facing AI systems and meeting the additional expectations for safety, trust, and polish.
Exposure to agent-driven and AI-generated user interfaces.

Responsibilities

Take concepts and business problems from rapid prototype through to reliable, scalable, production-grade systems.
Design, implement, deploy, and evolve autonomous and tool-using AI systems for internal and customer-facing use, owning their lifecycle and outcomes.
Design agent-orchestration architectures, including single- and multi-agent systems, tool and function calling, MCP and agent-to-agent coordination, planning and reasoning loops, memory, identity and access, and safety controls.
Assess problems individually and select appropriate models, agent patterns, and architectures.
Track emerging models, agentic frameworks, techniques, and tools; prototype promising options and bring suitable ones into production.
Deliver production software on time and to a high standard while meeting regulated-environment controls, governance, and compliance expectations.
Work closely with business stakeholders, shape the roadmap, and translate business goals into working AI capabilities.
Define standards for agentic AI quality, evaluation, observability, and responsible AI, and serve as a technical authority for AI delivery across the organisation.
Demonstrate the Senior Leader LEAD behaviours: Listen and be authentic, Energise and inspire, Align across the enterprise, and Develop others.
Demonstrate the Barclays Values of Respect, Integrity, Service, Excellence, and Stewardship, and the Barclays Mindset of Empower, Challenge, and Drive.