Nymbleup- AI Software Engineer - Backend Development
Location- Mumbai-Vikhroli
About the job : AI Software Engineer - Backend Development
We need a versatile engineer - someone who's just as comfortable shipping a backend service as chasing a frontend bug, writing a quick CLI, or prototyping an AI agent.
The work shifts week to week, and we need someone who finds that energising rather than chaotic.
If you already use Claude Code (or similar agentic tools) as part of how you actually work - not as a curiosity - and you genuinely enjoy debugging Linux systems when things go sideways, we should talk.
What you'll be doing :
Working across the stack - backend, frontend, infra, internal tools, AI features
Using Claude Code and agentic dev tools to ship faster, with the judgement to know when to lean on them and when not to
Debugging real problems across services, databases, browsers, and Linux hosts
Owning your code through design, ship, monitor, and on-call - not handing it off
What we're looking for :
- Bachelor's or higher in Computer Science / IT / Computer Engineering - this is a hard requirement. We are NOT considering candidates from non-CS streams (Mechanical, Electrical, ECE, Civil, Chemical, etc.) for this role.
- 3+ years of production Backend experience with at least one of : Python (Django/FastAPI), Ruby on Rails, Golang, or C++ - plus solid systems fundamentals (DSA concurrency, networking, OS)
- Hands-on familiarity with Claude Code (Cursor agents, Aider, Codex CLI, etc. count too)
- Strong Linux fundamentals - shell, processes, networking, systemd, the works
- Real debugging instincts - stack traces, logs, strace/tcpdump/journalctl, browser devtools
- PostgreSQL fluency
- Able to work full-time from our office
Bonus points if you have :
- Experience in QSR, retail, or similar high-volume operational domains
- Background in linear programming, optimisation, or demand forecasting
- Worked with LLM APIs (Anthropic, OpenAI) - tool use, evals, agentic workflows
- Docker, GCP/AWS, Sentry/Grafana
- Open-source or internal dev-tooling contributions
