Sr. Python Engineer, Agentic AI

Comcast Comcast · Media · Washington, DC

Senior Backend Engineer to lead technical direction of AI Agent initiatives, focusing on building scalable, reliable backend systems and agentic workflows. Responsibilities include architectural leadership, agent behavior engineering, system quality, testing, mentorship, and production ownership for customer-facing AI experiences.

What you'd actually do

  1. Design and own scalable backend architectures and complex agent-driven workflows. This includes service boundaries, data flows, failure handling, observability, and performance characteristics—not just model interaction.
  2. Design and productionize agent behaviors including tool invocation, state management, memory strategies, and reasoning control. Prompts are treated as code: versioned, tested, reviewed, and evolved alongside backend logic for reliability, cost, and latency.
  3. Set the bar for clean, readable, maintainable code across the Python backend. Establish patterns and standards that allow AI-heavy systems to remain understandable and operable over time.
  4. Design testing strategies for non-deterministic AI components, including behavioral regression testing, automated evaluations, and guardrails that reduce blast radius in production.
  5. Lead code reviews as a teaching tool, run architectural discussions, and actively mentor Engineer 1–2 developers in backend fundamentals, system design, and operational thinking.

Skills

Required

  • Python
  • backend systems design
  • distributed service architectures
  • asynchronous frameworks (FastAPI, asyncio)
  • agent behavior engineering
  • tool invocation
  • state management
  • memory strategies
  • reasoning control
  • testing strategies for non-deterministic AI components
  • automated evaluations
  • guardrails
  • CI/CD
  • deployment
  • runtime health
  • production support
  • on-call culture
  • microservices
  • data stores
  • queues
  • high-throughput APIs
  • mentorship

Nice to have

  • Critical Thinking
  • Curious Mindset
  • Customer-Oriented
  • Adaptability

What the JD emphasized

  • production-grade backend services
  • agentic workflows
  • customer-facing systems
  • agent behaviors
  • tool invocation
  • state management
  • memory strategies
  • reasoning control
  • backend logic
  • testing strategies
  • automated evaluations
  • guardrails
  • production support
  • customer-facing experiences
  • backend systems in Python
  • distributed service architectures
  • customer-facing systems in production
  • AI Systems Intuition
  • integrating LLMs into real systems
  • operational risks
  • workflows
  • control flow
  • safeguards
  • AI agents that interact with tools
  • APIs
  • external systems
  • managing agent state
  • bounding autonomy
  • preventing pathological behaviors
  • systems after launch
  • microservices
  • data stores
  • queues
  • high-throughput APIs
  • Mentorship & Engineering Culture
  • Pragmatism
  • stability
  • debuggability
  • long-term maintainability

Other signals

  • building scalable, reliable systems that happen to use modern AI techniques
  • architect and operate production-grade backend services and agentic workflows
  • designing for correctness, observability, failure recovery, and long-term maintainability
  • translate AI concepts into dependable, customer-facing systems