Principal Architect / Senior Staff Engineer (Agentic AI) | 9–15 Years | Bengaluru

Neuron7 is on a mission to redefine service excellence through Resolution Intelligence. We are seeking a Senior Staff Engineer for our Agentic AI team—a technical lighthouse and force multiplier who will architect the "Smart Resolution Hub" for the world's most complex service organizations. This is a role for a core builder who believes that "AI agents are software, not magic strings," and who has the engineering discipline to move beyond chatbots to create truly autonomous, goal-oriented systems.

• →Lead Agentic Architecture: Design and operate scalable multi-agent systems using planner-executor patterns and deterministic flow control.
• →Architect Memory at Scale: Build the hierarchical memory stack (Ephemeral, Semantic, Structured) that ensures Neuro—our next-gen agent—can maintain context and learn from past outcomes across thousands of product lines.
• →Data Intelligence Excellence: Lead the modeling of our Knowledge Graph infrastructure to unlock tribal knowledge and enable GraphRAG pipelines that reduce hallucinations to zero in mission-critical environments.
• →Platform-as-a-Product: Build the foundational Agentic Platform—including tool registries, identity-resolved profile services, and standard A2A protocols—that enables teams across Neuron7 to deploy agents with high velocity.
• →Inference Optimization: Take ownership of production readiness end-to-end, implementing intelligent model routing, semantic caching, and streaming to meet strict p95 latency targets (<2s).
• →AgentOps Leadership: Establish architectural standards for continuous evaluation and trajectory analysis. Mentor our engineers on how to build for failure, implementing circuit breakers and human-in-the-loop escalation paths for high-stakes decisions.
• →Continuous Adaptation: Lead our PEFT strategy, using LoRA-based fine-tuning to adapt open-source LLMs to niche service terminologies and deploying them efficiently via multi-adapter serving.

• Education: Master’s or PhD in Computer Science or a related technical discipline.
• Tenure and Impact: A minimum of 10-12 years of industry experience, with at least 3-5 years delivering production-grade AI/ML or Generative AI solutions at scale.
• System Mastery: Proven track record of architecting multi-agent orchestration systems (using LangGraph, AutoGen, or PydanticAI) and building complex, hierarchical memory stacks.
• Programming: High proficiency in Python and at least one compiled language (Java, Go, Rust, or C++) for performance-critical components.
• Data Intelligence: Expertise in Knowledge Graph modeling (Neo4j, Memgraph) and the construction of GraphRAG pipelines that outperform standard naive RAG in accuracy and explainability.
• Operational Discipline: Mastery of the full LLMops/Agentops lifecycle, including distributed tracing (OpenTelemetry), trajectory evaluation, and cost-per-goal optimization.
• Adaptation: Hands-on experience with parameter-efficient fine-tuning (LoRA, QLoRA) and multi-adapter serving frameworks.

• System Thinking: You clarify the problem before proposing a solution, reasoning through trade-offs in structured versus unstructured knowledge retrieval.
• Platform Intuition: You understand what it takes to build a "Golden Path" for other developers, moving from ad-hoc scripts to reusable platform modules.
• Direction Setting: You don't just solve the ticket; you identify which problems are worth solving to move the company’s EBITDA.
• Navigating Ambiguity: You thrive in the "fog," turning a vague requirement like "our diagnostic engine is slow" into a concrete multi-agent strategy.
• Influence without Authority: You can convince senior leadership and diverse engineering teams to follow your vision through data-driven prototypes and technical clarity.
• Taste: You possess the ability to distinguish between fragile prompt hacks and clean, extensible architectural solutions.
• Behavioral Agility: You are a lifelong mentor and unlearner. You have the courage to "let go of your legos" and let others do the work while you focus on the broader blast radius of your impact.