VERIZON AI PLATFORM

Forward Deployed Applied AI Engineer

Level: Senior Individual Contributor (Band 5–6)

Reports To: Head of AI Platform Engineering (dotted line to BU Tech Lead)

Location: Chennai,

Experience: 5–8 years engineering; 2+ years production LLM/GenAI; 1+ year agentic systems

The Role

You sit inside a Verizon business unit — not on the platform team — and turn the Verizon AI Platform's building blocks into production agents, workflows, and integrations that solve real operational problems. You discover the opportunity, build the solution, ship it, and measure it. You're also the platform team's eyes and ears: feeding signal back on what to build next.

What You'll Do

Solution Discovery & Agent Development

• Embed with domain teams to surface high-value AI opportunities and validate feasibility
• Design and deploy AI agents using LangGraph, Google ADK, or Verizon Agent SDK (pyvegas)
• Register agents in the Agent Registry with proper versioning, permissions, and MCP integrations

Context Engineering & Integration

• Build retrieval pipelines grounded in Customer 360, network telemetry, and product catalog data
• Wire agents to internal systems via the MCP Tool Gateway (billing, CRM, network inventory, Catalyst workflows)
• Handle auth, rate limiting, error recovery, and fallback logic — demo to production, not just demo

Evaluation, Adoption & Feedback

• Define success metrics with the BU; build eval harnesses and continuous benchmarking
• Tune guardrails for domain risks (CPNI, safety-critical network ops, financial accuracy)
• Run workshops, pair-program with domain engineers, and turn AI skeptics into AI builders
• File platform improvement tickets — you're the voice of the BU to the platform team

What You Won't Do

You don't build platform infrastructure, set RAI policy, select models, or manage a team. You're a force-multiplier IC focused entirely on shipping domain solutions that work at scale.

Required Qualifications

Technical — Must Have

• Python: production-grade (not notebook-grade)
• Agentic frameworks: LangGraph, Google ADK, or CrewAI — multi-step agents with tool use, memory, and error recovery
• LLM integration: GPT-4, Claude, or Gemini APIs; prompt engineering, structured outputs, function calling
• RAG pipelines: embeddings, vector stores, reranking, hybrid search — designed and deployed in production
• API & systems: REST/gRPC, OAuth/SAML, message queues, event-driven architecture
• Evaluation & observability: eval harnesses, trace analysis, quality monitoring for AI systems
• MCP / Tool Orchestration: model context protocol, tool registries, schema-driven tool dispatch

Technical — Strong Plus

• Verizon Agent SDK (pyvegas) or similar enterprise agent SDK
• Kubernetes / containerized deployment for AI workloads
• Streaming data (Kafka, Flink) for real-time context enrichment
• Fine-tuning, RLHF, or RLAIF experience

Non-Technical — Equally Critical

• understand unit economics — handle time, deflection rate, MTTR, NPS, conversion Business acumen:
• uncover the real problem, not just fulfill the stated request Consultative mindset:
• translate AI trade-offs for non-technical stakeholders; present to VPs Communication:
• measured on agents in production, not on decks presented Bias to shipping:
• you discover the problem, define the approach, build it, and measure it Comfort with ambiguity:

Reach me at suman@beaninfosys.com / +91 9381175498

Pay: ₹2,500,000.00 - ₹3,000,000.00 per year

Work Location: In person