Bring enterprise-grade GenAI to regulated healthcare workflows with CGI Group. This hybrid role in Lafayette, LA combines hands-on engineering with architecture governance to design and scale agentic AI capabilities for a Health Care client. You’ll work alongside the AI Center of Excellence (AI CoE) to align with enterprise AI standards and drive adoption across delivery teams.
Compensation: $80,600 - $218,200 per year (U.S. range estimate).
Experience level: 12+ years.
What you’ll do
You will own the architecture and technical direction for enterprise agentic AI solutions, balancing design governance with validation through working code. The goal is to establish reusable patterns that improve reliability and consistency across multi-agent orchestration, MCP-enabled tool ecosystems, secure backend services, integration, observability, and production readiness.
- Architect and implement robust multi-agent orchestration using Python and LangGraph, including intelligent routing, dynamic handoffs, shared state management, long-term memory, tool calling, error recovery, and follow-up conversation flows.
- Define target and reference architectures for agentic AI, retrieval augmented generation (RAG), model access, memory, evaluation, and human in the loop controls.
- Establish reusable standards, guardrails, and design review practices for enterprise AI delivery, including engineering patterns for security, maintainability, and reliability.
- Design, develop, and maintain MCP servers using FastAPI and Python to expose tools, resources, prompts, and custom capabilities within the agent ecosystem.
- Define and operationalize tool contracts covering discovery, versioning, permissions, schema validation, error handling, and safe execution.
- Build and maintain custom tools and integrations connecting AI agents with internal APIs, enterprise data sources, and legacy systems.
- Develop complex AI-powered chatbots and insights engines for healthcare and pharmacy benefit management workflows such as claims, pharmacy search, drug coverage, and prior authorization.
- Select model, RAG, memory, orchestration, and tool use approaches based on quality, latency, cost, security, and compliance needs.
- Own high-performance FastAPI services including streaming responses, asynchronous processing, custom middleware, rate limiting, caching, and authentication/authorization strategies.
- Design resilient integration patterns for APIs and legacy platforms, including retries, idempotency, timeouts, circuit breakers, fallbacks, and auditability.
- Ensure implementations follow best practices for security, privacy, responsible AI, and regulated data handling, including secrets management and input/output validation.
- Implement comprehensive observability with LangSmith, distributed tracing, monitoring, logging, evaluation, and performance tuning for AI agent workloads.
- Define service objectives and quality measures for latency, answer quality, tool accuracy, task completion rate, reliability, and cost.
- Identify and prioritize improvements in agent performance, latency, tool accuracy, scalability, and overall system architecture.
- Collaborate with frontend, DevOps, product, security, data, and enterprise architecture teams to deliver end-to-end AI capabilities with high reliability.
- Lead architecture reviews, mentor developers, communicate design tradeoffs, and provide guidance to distributed delivery teams.
- Evaluate emerging AI technologies pragmatically and recommend adoption based on measurable business value and enterprise readiness.
- Maintain end-to-end ownership from discovery and design through implementation, production readiness, and adoption.
- Drive architectural simplification, reuse, and consistent engineering practices across AI implementations.
- Act as a hands-on technical leader by validating critical designs and patterns through working code and prototypes.
- Build trusted partnerships across cross-functional teams to support delivery and adoption.
What you bring
- 12+ years of professional software engineering and solution architecture experience, including hands-on delivery of enterprise applications and platforms.
- 5+ years of AI/ML experience with recent hands-on delivery of production Generative AI and agentic AI solutions.
- Expert-level Python skills and experience building scalable, secure, production-grade backend services using FastAPI.
- Hands-on experience with LangGraph (or a comparable framework) for multi-agent orchestration including routing, handoffs, shared state, memory, tool use, recovery, and conversational continuity.
- Experience designing and implementing MCP servers, including governance of tools, resources, prompts, schemas, permissions, and lifecycle management.
- Deep understanding of LLMs, prompt engineering, RAG, embeddings, vector search, memory patterns, function/tool calling, model evaluation, and guardrails.
- Proven experience building AI-powered chatbots, insights engines, or workflow automation that integrates with enterprise APIs, data sources, and legacy systems.
- Strong knowledge of API and distributed system patterns, including streaming, asynchronous processing, middleware, rate limiting, authentication, authorization, resilience, and observability.
- Experience implementing LangSmith (or equivalent) for observability, tracing, evaluation, debugging, and performance monitoring.
- Strong understanding of cloud-native architecture, containers, CI/CD, security, privacy, responsible AI, and operational support for regulated enterprise workloads.
- Demonstrated ability to lead architecture decisions, mentor teams, facilitate design reviews, and communicate complex tradeoffs to technical and business stakeholders.
Technologies
- Python
- Large Language Models (LLMs)
- Prompt engineering
- Embeddings, vector databases, vector search
- Agent-based architectures
- LangGraph
- MCP servers
- FastAPI
- LangSmith
- Retrieval augmented generation (RAG)
- Function/tool calling
CGI benefits
- Competitive compensation
- Comprehensive insurance options
- Matching contributions through the 401(k) plan and the share purchase plan
- Paid time off for vacation, holidays, and sick time
- Paid parental leave
- Learning opportunities and tuition assistance
- Wellness and Wellbeing programs
Work location: Hybrid model with selected client locations including Bloomfield, CT, Raleigh, NC, or Lafayette, LA.
