The Company, a USA Subsidiary is a rapidly growing, private equity-backed SaaS company founded by engineers, focused on building scalable, high-quality products. Our solutions support over 3000 organizations, enabling them to manage grants, scholarships and philanthropic initiatives effectively. We offer cloud-based platforms that power the end-to-end lifecycle of grants, scholarships, fellowships, employee giving, and volunteer programs. In our organization, we foster a collaborative, innovation-driven culture with a flexible work environment and competitive benefits.
About the role
We're building the AI layer that makes our grant management platform - and the people who build it -more leveraged. We've already shipped the foundations: an internal MCP server with a growing tool catalog, a shared context layer that grounds agents in real product data, and a QA AIification initiative that's moving testing from local dev into staged release. We need an engineer to push this work forward
end-to-end.
You'll design, build, and operate agentic systems and AI-powered automations that touch real production workflows - both internal (engineering, QA, support tooling) and customer-facing (workflow rules, document processing, intelligent assistance inside the product). This is a hands-on builder role for someone who's excited about LLMs as a serious engineering substrate, not a demo.
Responsibilities:
- Design, build, and maintain MCP servers and tools that expose our internal systems (MongoDB, GraphQL APIs, internal services) to LLM agents safely and usefully
- Own AI-powered automations across the SDLC: spec-to-code workflows, automated PR review, QA generation, release-time checks
- Build customer-facing AI features - for example, replacing complex DSL-based rule engines with plain-English LLM-driven workflow creation
- Develop the shared context layer: retrieval, grounding, prompt assembly, and the evaluation harness that keeps it honest
- Implement evals, regression tests, and observability for LLM systems - latency, cost, accuracy, hallucination rate, drift
- Partner with product engineering teams to integrate AI capabilities into the Node.js / React/GraphQL / MongoDB / Apollo Federation stack
- Set patterns and guardrails: prompt management, tool security, rate limits, dry-run/safety modes for destructive operations
- Stay close to the frontier - evaluate new models, frameworks, and patterns as they emerge and bring the useful ones in
How we will take care of you:
- Motivating compensation
- Medical & Life insurance
- Paid Holidays
- Great working environment
- Rapid career development opportunities
Requirements
Must Haves:
- 4+ years of backend or full-stack engineering experience, with at least 1 year focused on LLM applications, agents, or AI-powered automation in production
- Strong proficiency with Python and/or TypeScript/Node.js
- Hands-on experience with the LLM application stack: OpenAI/Anthropic APIs, function calling/tool use, structured outputs, streaming
- Familiarity with MCP (Model Context Protocol) or comparable agent-tooling protocols and frameworks (LangGraph, LlamaIndex, custom orchestration)
- Practical experience with retrieval (vector or hybrid), prompt engineering, and LLM evaluation -you know how to make these systems reliable, not just functional- Solid software engineering fundamentals: testing, observability, code review, incident response
- Comfort with MongoDB or similar NoSQL databases, and with REST/GraphQL APIs
Nice to haves:
- Experience operating Claude Code, Cursor, or similar agentic coding tools at team or org scale
- Background in QA automation, test generation, or developer productivity tooling
- Familiarity with Apollo Federation, GraphQL subgraph architecture
- Experience with workflow/rules engines, DSLs, or no-code platforms
- Track record of shipping AI features that customers (not just internal users) actually adopt
- Familiarity with Playwright, Bruno, mabl, k6, or similar testing tools
How you'll be measured:
- AI capabilities shipped to internal teams and customers - adoption, reliability, retention of use
- Engineering velocity unlocked across the org through automation
- Operational health of AI systems: cost per outcome, accuracy/eval scores, incident rate
- Quality of the patterns and primitives you establish for others to build on
Your email job alert is now active!