Senior GenAI​/AI Solutions Architect

Work Schedule
Standard (Mon-Fri)

Environmental Conditions
Office

Job Description
As part of the Thermo Fisher Scientific team, you’ll discover meaningful work that makes a positive impact on a global scale. Join our colleagues in bringing our Mission to life every single day to enable our customers to make the world healthier, cleaner and safer. We provide our global teams with the resources needed to achieve individual career goals while helping to take science a step beyond by developing solutions for some of the world’s toughest challenges, like protecting the environment, making sure our food is safe or helping find cures for cancer.

DESCRIPTION:

We are seeking a  Senior/Lead GenAI/   AI Solution Architect  to translate business challenges into scalable, secure, enterprise-grade AI solutions across the  PSG value chain  (Commercial Operations, Finance/Legal, Manufacturing, Quality, Supply Chain). This role sits within  IT / Enterprise Architecture  and partners with PSG business teams, Engineering, Data/Platform, Security, and Quality/Validation to take problems from  concept → POC → prototype validation → enterprise deployment  , operating within  GxP expectations  where applicable.

KEY RESPONSIBILITIES:

Business to Solution Delivery:  Partner with stakeholders to define problem statements, success metrics, and solution concepts; deliver end-to-end implementations from  POC to scaled enterprise deployment  (security, governance, reliability, cost).

GenAI / Agentic Architecture:  Design and build advanced GenAI applications including  RAG, agentic RAG, and multi-agent orchestration  , integrating into existing enterprise systems and workflows.

Hands-on Prototyping & Implementation:  Build working solutions from the ground up in  Python  (services/APIs, integrations, testing, and telemetry) to demonstrate value quickly; iterate with users to validate usability and outcomes, then harden for production.

Enterprise Agent Tooling & Standards:  Establish reusable patterns, reference architectures, and organizational standards for agentic systems (e.g., internal enablement artifacts such as skills/standards documentation, and agent/tooling foundations).

LLM Evaluation & Quality Gates:  Implement evaluation frameworks (automated + human-in-the-loop) for retrieval quality, groundedness, accuracy, safety, latency, and cost; set up regression testing for prompts and workflows.

Prompt & Context Engineering:  Own best practices for prompt and context engineering (tool schemas, prompt/version management, context construction, retrieval tuning, and guardrails).

Agent Interoperability Patterns:  Implement agent interoperability patterns (e.g.,  Model Context Protocol (MCP)  and  agent-to-agent messaging patterns  ) in enterprise contexts (message contracts, routing, auditability, and boundaries).

Platform & Integration:  Work across  AWS  , OpenAI services,  Databricks  ,  Dataiku  , and  SQL  ecosystems to enable data access, orchestration, deployment, and monitoring.

GxP/Validation Partnership (as applicable):  Partner with Quality/Validation and Security to support required documentation, controls, and traceability for regulated or quality-critical deployments.

MINIMUM QUALIFICATIONS:

7+ years in solution architecture and/or senior engineering roles delivering enterprise systems.

2+ years of experience delivering Generative AI solutions in an enterprise environment, including taking solutions from prototype to production-scale deployment.

Demonstrated ability to take a business problem through  solution concept → POC → prototype validation → enterprise scale  .

Demonstrated product and outcome orientation, with the ability to prioritize work based on measurable business impact and end-user value.

Strong hands‑on  Python  experience delivering GenAI systems in an enterprise environment (building services/APIs, integrations, tests, and telemetry).

Practical experience with  Lang Chain, Lang Graph, and Lang Smith  (tracing, debugging, evaluation, and/or prompt/workflow regression).

Experience implementing  LLM evaluation frameworks  and measurable quality gates for  RAG/agentic workflows  (automated testing + human review…