Sr. Software Engineer; Remote

Jop Summary

You’ll be one of the first engineers on Neumo’s AI team, building the infrastructure and agentic systems that power AI adoption across the entire organization. The work spans three areas:

You’ll build and maintain the engines that make AI work at Neumo: our implementations engine, AI‑augmented SDLC tooling, agentic workflow orchestration (n8n, Lang Graph, Open Claw), and integration with our Platform services (MCP, LLM Service, RAG). You’re building the shared foundation that every team at Neumo will eventually consume.

You’ll drop into business processes, quickly understand the workflow, and build effective AI‑powered skills and automations. Then you’ll hand those skills off to non‑technical people so they can maintain and extend them without engineering involvement. The goal is to multiply what everyone at the company can do, not to become a permanent bottleneck.

AI moves fast. You’ll be expected to continuously test new tools, models, frameworks, and patterns, then bring what works back to the team. You’re the person who knows what’s possible today and what’s about to be possible next quarter.

• Design and implement safety, isolation, and governance controls for agents running in production
• Build orchestration pipelines using n8n, Lang Graph, or similar frameworks
• Create reusable automation patterns that other teams can adopt without your direct involvement
• Rapidly prototype AI solutions for business processes (support triage, document processing, configuration automation, SDLC workflows)
• Write effective prompts and prompt chains for complex multi‑step tasks
• Optimize for cost, latency, and accuracy across LLM providers
• Contribute to the AI‑augmented SDLC: coding assistants, automated testing, deployment automation, ticket quality validation
• Hand off completed skills and automations to non‑technical operators with clear documentation
• Perform other duties as assigned

• 3+ years of software engineering experience, building and shipping production systems
• A degree in AI, ML, or data science. We care about what you’ve built.
• Hands‑on experience building applications that integrate with LLMs (prompt engineering, API integration, structured output parsing)
• Experience with agentic patterns: tool use, multi‑step reasoning, human‑in‑the‑loop workflows
• Experience with cloud platforms (AWS preferred)
• Experience with agentic orchestration frameworks (Lang Graph, n8n, CrewAI, or similar)
• Experience with RAG architectures (vector databases, document chunking, retrieval optimization)
• Experience with Claude, Cursor, or similar AI‑assisted development tools
• Experience building systems that non‑technical users operate and maintain
• Experience mapping end‑to‑end business processes and identifying which steps can be automated vs. which require human judgment

• Strong fundamentals in at least one backend language (Node.js preferred, Python also valued)
• Comfort working across the stack when needed (APIs, databases, cloud infrastructure, basic frontend)
• Strong communication skills, both written and verbal. You’ll work with non‑technical stakeholders regularly.
• Ability to learn new tools and frameworks quickly. The landscape changes monthly.
• Familiarity with MCP (Model Context Protocol) or similar LLM connectivity patterns
• Understanding of LLM cost optimization (model selection, caching, prompt efficiency)
• ML/model training experience. We use LLMs as services, not train them.
• Government or gov‑tech experience. Helpful context, but not a filter.

• Agentic‑first, not chatbot‑first. We build agents with human‑in‑the‑loop verification, not chat interfaces.
• Centralized agent development. All new agent work runs through this team until risk and patterns are well understood. You own build, ops, and governance.
• Shared services model. You build on Platform’s foundation (MCP, LLM Service, RAG) and create reusable patterns. No one‑off implementations that don’t generalize.
• Speed matters. We’re in a 1‑year competitive window. Bias for action, rapid prototyping, and fast iteration.
• Tools:
  Claude, Type Script,…