Senior AI/ML Architect
Role Overview
In this position, the Senior AI/ML Architect acts as the go-to expert for the agentic AI track across enterprise engagements. The work centers on running client-facing discussions to surface and frame use cases, shaping agentic AI designs, and steering their build-out on AWS Bedrock AgentCore and the services around it. The role pairs hands-on AI/ML engineering with the credibility to serve as a technical advisor to stakeholders on the client side, and it carries ownership of enterprise-scale AI/ML delivery end to end — the kind that automates high-volume, complex workflows for large organizations.
Success here calls for real technical range spanning AI/ML and generative AI, a knack for ramping quickly on whatever industry an engagement lands in, and the program leadership needed to push transformation forward at scale.
What the Day Looks Like
Most days open with facilitating technical working sessions and client conversations aimed at spotting, ranking, and scoping candidate agentic AI use cases. Expect to operate shoulder to shoulder with mixed teams — data scientists, AI/ML engineers, cloud architects, business analysts, and subject-matter experts on the client's side — turning business challenges into agentic patterns and solutions that are ready for production.
The rhythm moves back and forth between the advisory side — guiding client stakeholders as their trusted technical resource — and the builder side: sketching RAG architectures, standing up agents on AWS Bedrock AgentCore, and wiring the evaluation pipelines that hold agentic systems to standards for accuracy, safety, speed, and spend. Every architect on this team is an engineer at the core, fluent across the whole delivery lifecycle and accountable for their solutions from early roadmap straight through to production.
Core Responsibilities
- Map agentic patterns — ReAct, tool-use, and multi-agent orchestration — onto the specific business problems a client brings to the table
- Stand up agents with an agent framework (LangChain / LangGraph, Strands Agents, or comparable) running on Bedrock AgentCore Runtime
- Own the AI/ML solution roadmap for full-lifecycle process automation, keeping it tied to where the client wants its enterprise transformation to go
- Facilitate technical workshops with clients to surface, prioritize, and scope agentic AI opportunities
- Build RAG architectures on Bedrock Knowledge Bases paired with vector stores
- Set the AI/MLOps direction, covering model monitoring, retraining pipelines, drift detection, and test coverage
- Partner across disciplines — data scientists, AI/ML engineers, cloud architects, business analysts, and client-side SMEs — to ship production-grade AI/ML work
- Instrument agents for tracing, monitoring, and troubleshooting through Bedrock AgentCore Observability
- Design AI/ML pipelines that span the full path: data intake and extraction, automated decisioning, validation, and the downstream processing that follows
- Stand up and operate evaluation pipelines that keep agentic systems honest on accuracy, safety, latency, and cost
- Guide model work across the lifecycle — development, training, validation, and deployment
- Engineer for scale so solutions absorb enterprise volumes reaching into the millions of transactions each month
Required Background
- A decade or more in technology delivery, including five-plus years at the helm of enterprise-scale AI/ML programs
- A track record as an AI/ML subject-matter expert, ideally earned inside a consulting or professional services setting
- Hands-on depth building and deploying agentic AI solutions and the patterns behind them (ReAct, tool-use, multi-agent orchestration)
- Working experience across AWS AI/ML and generative AI services — Amazon Bedrock, Bedrock AgentCore, and Bedrock Knowledge Bases among them
- Practical experience architecting RAG solutions backed by vector stores
- Substantial hands-on AI/ML engineering chops, including taking models through development, training, validation, and production deployment
- Familiarity with AI/MLOps practice — model monitoring, retraining pipelines, drift detection, and testing
- Sharp communicator and collaborator who can hold the trusted-advisor seat with client stakeholders and steer client-facing workshops
- Strong program leadership, with a talent for aligning cross-functional teams and driving change at enterprise scale
- Solid analytical and problem-solving instincts, translating business needs into workable technical designs
- Comfort ramping fast on unfamiliar industries and business processes as engagements shift
- Fluency navigating ambiguity — shaping vague client wants into stories and epics a team can actually execute inside a sprint (in other words, at home in an agile delivery cadence)
- A self-starter who stays curious and keeps pace with the fast-moving agentic AI space
Technical Toolkit
- Python
- Amazon Bedrock, including the Bedrock AgentCore Runtime and Observability components
- Agent development frameworks such as LangChain / LangGraph or Strands Agents (or a close equivalent)
- Agentic AI design patterns: ReAct, tool-use, and multi-agent orchestration
- Embedding generation together with vector storage and retrieval methods and tooling
Nice to Have
- Exposure to the healthcare space — payer systems, or claims work spanning eligibility, intake, adjudication, denial management, and payment
- AWS Certified Generative AI Developer – Professional, or comparable AWS credentials