Leverage your problem-solving skills to thrive in a fast-paced environment and drive customer-centric strategies. As a leader in solutioning, collaborate closely with the Sales teams to deliver tailor-made product solutions that meet clients’ needs. Ignite your passion for product innovation by leading customer-centric development, inspiring solutions, and shaping the future with your strategic vision and influence. You will lead innovation through the development of products and features that delight customers. As a leader on the team, you leverage your advanced capabilities to challenge traditional approaches, remove barriers to success, and foster a culture of continuous innovation that helps inspire cross-functional teams create groundbreaking solutions that address customer needs.
As a Product Manager in Chief Data & Analytics Office - Fusion Platform Team, you are an integral part of a team that defines and configures complex solutions for key client relationships and prospect opportunities in partnership with Sales. You are responsible for acting as the voice of the customer by understanding their needs and communicating feedback to the Product teams. The Agent Platform Engineer role sits inside the Product Agent Solutions Team the Forward Deployed Engineering function within CDAO (Chief Data & Analytics Office). This team is the tip of the spear for agentic AI adoption across the firm. We partner directly with Line of Business engineering teams, get in the code with them, and do not leave until a production agent ships and if you want to build AI systems that actually run in production at one of the world's most complex regulated environments, this is the role.
As an Agent Platform Engineer, you are first and foremost a builder. You write code, design agent architectures, and solve the hard integration problems that stand between a working agent prototype and a production system serving internal clients in a regulated environment. You are not a Product Manager, not a deck builder, and not a facilitator - you are an engineer who can also think in systems and communicate with senior stakeholders. You will work within the Agent Builder ecosystem contributing to reference implementations, debugging real integration failures, and developing the reusable patterns that scale agent adoption across the firm. This role demands both breadth (platform fluency across the full agent stack) and depth (the ability to architect a multi-agent system and write the code to prove it works).
Job responsibilities
- Leads solutioning and the adoption of existing and upcoming client-facing products and capabilities while defining and configuring optimal solutions that address clients’ needs and objectives
- Serves as a subject matter expert on a defined set of products and capabilities with a deep understanding of our clients’ needs and current industry trends
- Supports Sales in pricing, pipeline planning, account planning, and upskilling the team on product knowledge by collaborating on training and collateral materials
- Engages with client teams to better understand pain points and refine solutions while regularly communicating critical client feedback to Product teams to inform the strategic product roadmap
- Design and build production-grade AI agents using Agent Studio, SmartSDK, RAG SDK, and MCP SDK including orchestrator/sub-agent architectures, tool-calling patterns, parallel execution loops, and write-back integrations.
- Partner directly with LoB (Line of Business) engineering teams in Forward Deployed Engineering engagements — embed alongside their engineers, debug live integration issues, and jointly ship production agents on Fusion.
- Architect multi-agent systems: define agent boundaries, orchestration patterns, context passing, tool surface exposure, and state management for regulated production workloads.
- Develop and maintain reference implementations and SDK playbooks that translate platform capabilities into reusable, opinionated engineering patterns for LoB (Line of Business) consumption.
- Contribute to MCP SDK design and tooling — define tool schemas, validate tool surface security, and build integrations between agents and enterprise systems. Integrate RAG pipelines into agent workflows — manage knowledge base configuration, chunking strategies, retrieval tuning, and drift monitoring in production.
- Identify and close capability gaps in agent observability, evaluation, and error recovery — work with Platform Engineering to surface and prioritize field-driven requirements. Participate in architecture reviews for high-complexity LoB (Line of Business) agent builds — provide hands-on guidance on blast radius containment, human oversight hooks, and production hardening.
- Contribute to the Agent Deployment Risk Framework — translate governance requirements into engineering constraints that ship as code, not documentation. Maintain personal technical depth as the agent stack evolves — MCP, tool-calling patterns, multi-modal inputs, model gateway integration, and evaluation frameworks.
Required qualifications, capabilities, and skills
- 5+ years of experience or equivalent expertise in problem-solving across multiple teams and a cluster of products
- Extensive experience working in a sales cycle and engaging with clients on a regular basis
- Experience modifying preconfigured solutions to meet complex problems
- Demonstrated prior experience working in a highly matrixed and complex organization
- 5+ years of software engineering experience, with at least 3 years focused on AI/ML systems, GenAI application development, or agent-based architectures in production.
- Strong Python fluency — you write production-quality Python, not just scripts. Experience with async patterns, SDK extension, and framework-level engineering is expected.
- Hands-on experience building agents or agentic workflows — tool-calling, orchestration, multi-step reasoning loops, and agent-to-agent communication patterns. Working knowledge of LLM APIs and agent frameworks (LangChain, LangGraph, AutoGen, CrewAI, or equivalent) not just tutorials, but actual production systems.
- Experience integrating RAG pipelines: vector stores, embedding models, chunking strategy, retrieval evaluation, and production monitoring.
- Ability to architect systems at the component level — define interfaces, trace data flows, identify failure modes, and reason about blast radius in distributed agent systems.
- Comfortable operating in complex enterprise environments with governance, compliance, and model risk constraints — you understand why these exist and how to engineer around them, not just complain about them.
- Strong written and verbal communication - you can explain an agent architecture to a senior engineer and to a business MD, without changing the truth in between.
Preferred qualifications, capabilities, and skills
- Direct experience with MCP (Model Context Protocol) designing tool schemas, building MCP servers, managing tool surface exposure, or integrating MCP into an agent platform.
- Experience in regulated industries — financial services, healthcare, or government — with practical exposure to model risk management, audit trails, and compliance-driven engineering constraints.
- Familiarity with agent security concerns: prompt injection, tool misuse, over-privileged tool access, and blast radius containment strategies.
- Experience building evaluation frameworks for LLM-based systems not just benchmarks, but production-grade evaluation pipelines with structured outputs and regression tracking.
- Exposure to cloud-native AI infrastructure — managed model endpoints, model gateways, token/cost observability, and multi-tenant serving considerations.
- Experience contributing to developer-facing SDK or platform tooling — designing APIs that other engineers consume, writing documentation that actually gets used, and iterating based on adoption signal. Familiarity with responsible AI practices as they apply to agents: human oversight requirements, escalation paths, intervention hooks, and auditability standards.