Senior Engineer, Applied AI & Engineering Platforms

Company Description

About AbbVie

AbbVie's mission is to discover and deliver innovative medicines and solutions that solve serious health issues today and address the medical challenges of tomorrow. We strive to have a remarkable impact on people's lives across several key therapeutic areas including immunology, oncology and neuroscience - and products and services in our Allergan Aesthetics portfolio. For more information about AbbVie, please visit us at www.abbvie.com. Follow @abbvie on LinkedIn, Facebook, Instagram, X and YouTube.

Job Description

Join an inclusive, collaborative Business Technology Solutions (BTS) team as a Sr Engineer, Applied AI & Engineering Platforms at AbbVie. This is a hands-on, lead technical engineering role at the center of AbbVie’s generative and agentic AI transformation — building intelligent, autonomous systems and scalable agentic workflows that will accelerate drug discovery, streamline clinical and regulatory operations, and reimagine how AbbVie works across every function. 

You will design and own the AI foundations layer that underpins all agentic capabilities across the enterprise, establish engineering standards that make AI systems reliable and auditable in GxP-regulated environments, and serve as a technical authority guiding platform teams, data scientists, and application engineers across the organization. 

This is not a research or prototyping role. You will architect, build, and operate production-grade multi-agent systems used in clinical, commercial, and operational domains — working alongside enterprise architecture, platform security, data engineering, MLOps, and domain subject matter experts to ensure every system is deployable, governed, and compliant from day one. 

Responsibilities:

Agentic System Design & Engineering 

• Architect and own production-grade multi-agent systems using orchestration frameworks (LangChain, LangGraph, CrewAI, OpenAI Agents SDK, AutoGen/AG2, Semantic Kernel), making deliberate decisions on state management, routing, memory architecture, and failure handling. 
• Design agent cognitive architectures — planning loops (ReAct, Reflexion, CoT), tool-use patterns, memory systems (short-term, episodic, semantic), and self-evaluation loops. 
• Build multi-agent coordination patterns (supervisor–worker, peer collaboration, A2A protocols) aligned with emerging open standards including MCP server integration to connect agents to enterprise systems, clinical data platforms, and regulatory repositories. 

AI Foundations Layer 

• Design and maintain shared AI infrastructure: LLM gateway/routing, embedding services, vector stores, RAG pipelines, prompt management, and model evaluation harnesses across all agentic products. 
• Establish model selection and governance spanning hosted providers (Claude, GPT, Gemini) and self-hosted models, including fine-tuning pipelines (LoRA/QLoRA) for pharmaceutical-specific tasks. 
• Build context engineering standards — managing context windows, retrieval strategies, chunking, re-ranking, hybrid search, and query routing for enterprise-scale clinical and scientific knowledge — with guardrails, safety layers, content filters, and HITL escalation appropriate for GxP environments. 

Agentic Engineering SDLC 

• Define the end-to-end SDLC for agentic systems — from design through evaluation, deployment, and continuous monitoring — treating agent behavior as a first-class software artifact subject to change control. 
• Build agent evaluation frameworks (golden test sets, LLM-as-judge scoring, regression detection, task-completion benchmarks, latency/cost dashboards) and CI/CD pipelines with automated evaluation gates, drift detection, and rollback capabilities. 
• Establish traceability, audit logging, and versioning standards supporting GxP validation, 21 CFR Part 11, and AbbVie’s AI governance policy. 

Observability, Reliability & AIOps 

• Implement full-stack observability (LangSmith, Langfuse, OpenTelemetry): trace-level logging, token/cost tracking, latency profiling, and anomaly detection on agent behavior. 
• Own production reliability — retry logic, fallback strategies, circuit breakers, graceful degradation, and HITL escalation for regulated workflows. Monitor for behavior dr