Back End Developer

Founding Engineer — Backend, Data & Infrastructure Bangalore (Hybrid) or Remote within India · Full-Time · Reports to Founder/CTOWe're an early-stage B2B SaaS company. The MVP is built and running. You're the first full-time engineer, and you'll own everything from the data ingestion layer to what ships to production.This is not a role where you get handed tickets. You'll be deciding what gets built, how it gets built, and whether it's ready to go live.WHERE THINGS STAND TODAYThe core product logic works. The database schema is designed and running on PostgreSQL. The FastAPI backend is scaffolded with async architecture. CSV ingestion is functional. The frontend is deployed. What's missing is the layer between "it works on my machine" and "a paying customer depends on this." That's your mandate.You are not starting from zero, and you are not inheriting a legacy mess. You're inheriting a working prototype that needs an engineer who knows what production actually means.WHAT YOU'LL OWN▸ BACKEND & APITake the existing FastAPI + PostgreSQL codebase to production standard. Structured logging, proper error handling, a test suite that actually catches regressions, and monitoring that tells us when something breaks before a customer does. Build and maintain the REST APIs serving the frontend, with low-latency responses on table and drill-down queries.▸ DATA ENGINEERINGDesign and build the ETL layer. Ingest from third-party APIs and file uploads, transform, normalize into our core schema, and load reliably. Handle the unglamorous parts properly: schema drift, partial syncs, duplicate records, timezone mismatches, and data that arrives malformed because someone's CRM has a custom field nobody documented.▸ INTEGRATIONSOwn third-party connectors end to end — CRM, accounting, and ad platforms. OAuth flows, token refresh, rate limit handling, sync scheduling, retry logic, and failure recovery that doesn't require a human to notice. Every connector needs a CSV fallback path.▸ CORE ENGINEWork on the analytical logic at the heart of the product: time-series correlation, weighting, and confidence scoring. This is deterministic statistical work, not ML. You'll be tuning it for accuracy, hardening it against incomplete data and missing time windows, and building the test harness that proves it's right. Accuracy here is a commercial requirement, not a nice-to-have — we can't sell what we can't defend.▸ INFRASTRUCTURE & DEVOPSDocker, GitHub Actions CI/CD, AWS deployment, secrets management, backups, basic alerting. Keep it lean and boring. No Kubernetes until we have a real reason.▸ PRODUCT & PROCESSDaily syncs and weekly sprint planning with the founder. You'll have direct input on what we build and in what order. When you think a feature is wrong, say so — that's part of the job.▸ AS WE GROWEvolve the analytical models against real customer data. Move from overnight batch to near-real-time refresh. Build proper multi-tenancy. Mentor the interns now and the engineers we hire later.THE STACKBackend: Python 3.11+, FastAPI, PostgreSQL, Redis, Celery Data: Pandas / Polars, Airbyte or custom connectors Frontend: Next.js 14+, TypeScript, Tailwind, Recharts Infra: AWS, Docker, GitHub ActionsMUST HAVE2–4 years building production backend systems in PythonStrong PostgreSQL — schema design, complex queries, indexing, query performance tuning. You should be able to read an EXPLAIN plan and know what to do about itFastAPI (strongly preferred) or Flask/Django with async experienceReal ETL experience: you've pulled from third-party APIs, transformed the data, and loaded it somewhere that other people depended onPandas or Polars for data transformationRedis for caching and queueingDocker and CI/CDComfort with ambiguity, and the judgment to unblock yourself without waiting to be toldNICE TO HAVEIntegration experience with HubSpot, Salesforce, QuickBooks, Xero, or ad platform APIsCelery in productionAirbyte, Fivetran, dbt, or similarTime-series data, statistical correlation, or basic causal inferenceAWS deployment and infrastructure managementEnough React/Next.js to fix a frontend bug without escalatingA side project you built because you wanted to, not because someone askedWHAT THIS ROLE IS NOTThis is not a role with a defined backlog and a tech lead above you. If you need structure to be productive, this will be uncomfortable.You will be on call, informally, for a system you built. Early-stage means the person who wrote it is the person who fixes it at 11pm.The product will change. Some of what you build in month two will be deleted in month five. If that bothers you, this isn't the right stage.WHY TAKE ITEngineer #1. Every architectural decision from here is yours to make or argue against. As the team grows, this becomes a lead role by default, not by promotion cycle.Real users fast. Our sister agency serves the exact companies this product is for. What you ship gets tested against real operational data within weeks.Genuinely hard problem. Attribution across disconnected data sources — spend, headcount, vendor cost, revenue outcomes — is unsolved for good reasons. The interesting work is in the data quality and correlation layer, and there's no library that does it for you.Direct founder access. No layers, no committees, no six-month roadmap reviews.TO APPLYResume, plus a short note answering one question: what's the most interesting backend system or data pipeline you've built, what broke in production, and what you changed as a result?Applications without the note won't be read.