Years of Experience: 8+ years in software engineering
Job Mode: Full time
Work Mode: Remote within Mexico
The Agentic AI Expert (19071) will play a critical role in evaluating, refining, and advancing Broadcom’s Maverick Platform—an internal agentic AI pipeline designed to automate complex software engineering workflows. This role requires deep expertise in agentic systems, LLM orchestration, and reasoning models, with hands-on experience building and operating autonomous agents in Kubernetes-based environments.
Evaluate and improve Broadcom’s LangGraph-based workflow and reasoning engine. Review, refine, and optimize agent libraries, templates, and orchestration patterns. Assess and enhance agent autonomy, retries, memory usage, and context management. Analyze and improve agent identity, observability, safety, and governance mechanisms. Support agent-to-agent (A2A) communication patterns and execution workflows. Collaborate with platform teams to ensure scalable, reliable agent execution within Kubernetes pods. Provide technical guidance and best practices for expanding the agent ecosystem and roadmap.
Deep expertise in agentic AI systems and LLM-driven orchestration frameworks. Strong understanding of reasoning models and agent reasoning loops. Proven experience evaluating production-grade AI platforms and workflows. Ability to work independently and provide architectural-level recommendations. Strong communication skills to collaborate with distributed, cross-functional teams.
Hands-on experience with LangGraph or comparable agent orchestration frameworks. Strong understanding of MCP (Model Context Protocol) and managed identity patterns. Experience designing agents with memory, retries, observability, and safety controls. Demonstrated ability to build agents that run natively inside Kubernetes pods. Familiarity with CI/CD-driven agent execution and code automation workflows.
LangGraph, LLM-based reasoning models (Gemini or equivalent), MCP (Model Context Protocol), Kubernetes, Agent-to-Agent (A2A) communication, OAuth-based identity systems, GitHub, CI/CD pipelines, security and quality tools such as Blackduck and Fortify.
Agentic system architecture and design, LLM orchestration and workflow decomposition, autonomous agent execution and governance, observability, logging, and troubleshooting, secure and scalable cloud-native AI systems.
Encora is the preferred digital engineering and modernization partner of some of the world’s leading enterprises and digital native companies. With over 9,000 experts in 47+ offices and innovation labs worldwide, Encora’s technology practices include Product Engineering & Development, Cloud Services, Quality Engineering, DevSecOps, Data & Analytics, Digital Experience, Cybersecurity, and AI & LLM Engineering. At Encora, we hire professionals based solely on their skills and qualifications, and do not discriminate based on age, disability, religion, gender, sexual orientation, socioeconomic status, or nationality.