(USA) Staff, Software Engineer

Position Summary...

What you'll do...

About the Role:

We are looking for a Staff Software Engineer who is an expert in Python, with a deep mastery of LLM-driven systems, agentic architectures, and scalable intelligent automation. This is a pivotal role where you'll define and implement the technical strategy, influence architecture decisions, and lead the creation of next-gen AI-driven platform that move beyond prompt engineering into modular reasoning system.

This role isn’t about building one-off workflows—it’s about inventing and hardening intelligent systems that can reason, act, and adapt. You will shape the core architecture of multi-agent platforms, ensure LLM integrations are secure, efficient, and observable, and build frameworks that others can extend across use cases and orgs.

Key Responsibilities:

• Architect modular, testable, and composable Python systems that support multi-agent workflows, tool-chaining, RAG, memory management, and fallback strategies.

• Design LLM-powered execution engines that support both high throughput and adaptive reasoning (via LangChain, AutoGen, or custom frameworks).

• Lead implementation of retrieval-augmented generation (RAG) pipelines, semantic search, and structured knowledge memory systems.

• Build and scale integrations with internal LLMs, including handling signature-based auth, function calling, and context management at scale.

• Drive end-to-end lifecycle: from configuration schema (YAML) to execution trace logging, observability, and self-healing recovery patterns.

Must-Have Qualifications:

• 8 + years of professional software engineering experience, with 5+ years in Python, building distributed systems at scale.

• Deep knowledge of agentic design patterns, including:

• ReAct, Plan-and-Execute, AutoGen-style coordination

• Tool calling, dynamic agent routing, and recursive agent planning

• Semantic memory, embedding-based context lookup, summarization windows

• Expertise in building LLM-based systems with LangChain, OpenAI, Anthropic, or custom orchestrators.

Hands-on experience with:

• RAG pipelines using vector stores (FAISS, Pinecone, Weaviate, Qdrant, Azure Cognitive Search)

• LLM evaluation and observability (tracing, token usage, agent state tracking)

• Workflow orchestration using config-first approaches (YAML/JSON definitions, step runners, etc.)

• Proven ability to drive technical vision, resolve ambiguity, and make architectural tradeoffs at scale.

• Strong background in distributed systems, task queues, asynchronous workflows, and backend performance optimization.

• Experience in cloud-native environments (AWS, GCP, or Azure), including containerization, monitoring, and secure API integrations.

Nice-to-Have:

• Built or contributed to a custom agentic orchestration framework used across multiple product lines.

• Experience with vector search optimization, context ranking, or temporal memory solutions.

• Published talks, blogs, or papers on LLM systems, AI architecture, or applied reasoning frameworks.

• Deep understanding of how to apply LLM systems in regulated or high-compliance environments (PII handling, redaction, observability).

• Exposure to DevEx platforms for developers to build workflows on top of intelligent agents.

• Familiarity with multi-modal agents (text + vision), LLM simulation patterns, or offline evaluation loops.

What Success Looks Like:

You’ve built an agent platform that others across the org use as the foundation for intelligent automation.

You turn abstract ideas into clean, extensible, production-grade Python systems that scale and evolve.

You elevate technical conversations, coach Staff+ engineers, and are the go-to person for unblocking complex challenges.

You are not just LLM-aware—you pioneer how LLMs are applied in production systems, with a clear perspective on wh