Backend Product Engineer - Python

Product Engineer | Fibr.ai

Bengaluru - In office

Stack: Python - FastAPI - Temporal - PostgreSQL - MongoDB - Redis

The role:

Fibr is building the agentic web experience layer turning every URL into an intelligent agent that senses intent, makes decisions, and adapts in real time. Our agents personalise the full journey, running long LLM workflows against live customer accounts and millions of sessions.

You'll be a senior technical voice on the team shaping architecture, making the hard calls on system design, and owning the most complex parts of the platform. You'll work directly with the founding team and have real influence over where the product goes technically.

What you'll do:

• Own the hardest engineering problems durable workflows, eval infrastructure, agent reliability at scale

• Set the technical direction for new surface areas agent architectures, data models, integration patterns

• Ship AI-powered features end-to-end and hold the bar for quality across the team

• Design systems that survive rate limits, partial failures, flaky third-party APIs, and real production load

• Build the evals, guardrails, and instrumentation framework the whole team ships against

• Sit at the table for architecture decisions and drive them forward

• Mentor and raise the bar for the broader engineering team

Must-haves:

• 2-5 years of production backend experience

• Python is your primary language and we move fast

• Has shipped LLM-powered features to real users at scale not demos, not side projects. Talk about what broke, how you debugged it, and what you rebuilt

• Has designed and owned eval infrastructure for LLM systems show us the harness, the dataset, and how it drove ship/no-ship decisions

• Has taken systems from 1 to 100 not just 0 to 1. You've dealt with scale, reliability, and production incidents firsthand

• Deep experience with durable workflow systems in production Temporal, Celery, or similar at real scale

• Strong system design fundamentals you think about failure modes before you write the first line

• Strong product thinking alongside engineering depth you catch the product gap before the PM does

• Asks clarifying questions before jumping into solutions you understand the problem before you build the solution

• High ownership decisions are yours to make, not wait for

Nice-to-haves:

• Multi-agent architectures or complex tool-use pipelines in production

• Experience with rate-limited third-party APIs at scale ad platforms, CRMs, analytics tools

• Vector stores, RAG pipelines, or retrieval systems in production

• LLM observability and tracing LangSmith, Langfuse, Prometheus, or similar

• Time spent as a founding engineer or early engineer at a seed/Series A startup

• Has mentored junior engineers or led technical decisions in a small team

• Strong product sense you catch the system design issue and the UX issue

This role is not for you if:

• Your AI experience is limited to personal projects or hackathons

• You've only built demos or PoCs that never reached real users

• You need detailed specs before you can start building

• Your primary language is JavaScript or Java