Senior Software Engineer

• Internal LLM stack running on owned GPU infrastructure (~500 GPUs)
• GPU-as-a-Service on remaining capacity

Claude is integrated across the platform. We own the data plane, compute, and model layer - not just an API wrapper.

The role

We moved fast to get real customers live (PacketFabric, Qwilt, DTIQ, WeLink, Element8). The system works - but parts are still fragile.

Your job is to take what exists and harden it into production-grade infrastructure: reduce technical debt, improve reliability, and turn AI-generated scaffolding into a system a senior engineer would be proud to own.

What you’ll do

• Refactor the multitenant CRM/ticketing stack (Next.js, Cloud Run, Cloud SQL, Prisma) for correctness, performance, and isolation
• Stabilize and make deterministic the autonomous agent pipeline (bug triage * fix * review)
• Implement proper CI/CD, migration tracking, secret management, and deploy safety
• Build real observability (logs, traces, error budgets) instead of ad-hoc debugging
• Own integration between CRM, MCP servers, and LLM/GPU layers
• Set engineering standards while mentoring a mix of engineers and AI agents

Do you have experience in Systems engineering?, * 5+ years building production systems (SaaS, infra, or platform engineering)

• Strong TypeScript/Node + Postgres experience (Prisma a plus)
• Deep understanding of system design, failure modes, and data modeling
• Comfortable working with and improving AI-generated code
• Experience owning production systems and on-call responsibility
• GCP experience (Cloud Run, Cloud SQL, Secret Manager) is a plus
• Bonus: experience with LLM systems, agent workflows, or MCP