SEI is seeking an AI Architect. As part of the Enterprise AI team, you will be responsible for designing and delivering enterprise AI platforms-RAG systems, agentic frameworks, document intelligence services, and responsible-AI patterns-on Azure. You bridge strategy and execution: defining reference architectures, running fast POCs, writing production code, and partnering with InfoSec, Cloud, and business stakeholders.
What you will do:- Design AI architectures end-to-end: RAG pipelines, agentic orchestration (multi-agent, tool use, MCP), retrieval strategies, grounding, and structured outputs.
- Architect Document Intelligence as a Service: ingestion, OCR/vision, extraction, verification, multi-tenant design, and sensitive-data handling.
- Define enterprise patterns: prompt governance, model routing, security envelopes, HITL gates, plugin/MCP integration, fallback/verification flows.
- Prototype fast: build rapid POCs in Python to validate architecture, derisk decisions, and accelerate delivery.
- Drive TrustAI: hallucination testing, bias/fairness evals, lineage, scoring, drift monitoring, and compliance alignment.
- Design robust telemetry: retrieval/prompt tracing, eval feedback loops, guardrail instrumentation (OpenTelemetry, Langfuse, App Insights).
- Interface with InfoSec & Cloud Enablement: private networking, identity boundaries, policy-as-code, DLP, model-risk governance.
- Publish reference architectures, reusable modules, decision playbooks; lead design reviews and mentor engineers.
What we need from you: - A minimum of 10 years of experience with software/ML engineering; strong Python (APIs, services, production).
- Deep RAG expertise: embeddings, hybrid search, chunking, reranking, evals, guardrails.
- Agentic AI: multi-step agents, tool calling, orchestration frameworks (LangChain/LlamaIndex or custom).
- MCP / plugin patterns, tool-server integration, context protocol design.
- Azure: OpenAI/Models, AI Search, ML, AKS, Functions, Key Vault, VNET/private endpoints.
- Solid grasp of LLM lifecycle: prompting, structured outputs, token/cost management, fine-tuning concepts.
- Ability to design for regulated industries: PII handling, isolation, audit, governance readiness.
- Strong communication; able to present to C-suite, InfoSec, and engineering audiences.
What we would like from you:- Experience effectively interacting with large application development and delivery teams and application development managers who are providing full lifecycle application support in complex, heterogeneous environments
- Experience working with senior leadership, ability to work comfortably with a wide range of people and skill sets
- LLM fine-tuning (LoRA/QLoRA), evaluation frameworks (RAGAS, pairwise).
- Vector databases (Milvus, Pinecone, Weaviate).
- TrustAI implementation: safety evals, governance dashboards, drift monitoring.
- TypeScript/Node for plugins or front-end integrations.
This team follows a hybrid schedule and work 3 days in the office.
To help you stay energized, engaged and inspired, we offer a wide range of benefits including comprehensive care for your physical and mental well-being, a strong retirement plan, tuition reimbursement, a hybrid working environment for most roles, support for working parents and flexible Paid Time Off (PTO) so you can relax, recharge and be there for the people you care about.
Benefits include healthcare (medical, dental, vision, prescription, wellness, EAP, FSA), life and disability insurance (premiums paid for base coverage), 401(k) match, education assistance, commuter benefits, up to 11 paid holidays/year, 21 days PTO/year pro-rated for new hires which increases over time, paid parental leave, back-up childcare arrangements, paid volunteer days, a discounted stock purchase plan, investment options, access to thriving employee networks and more.
