Founding Staff Engineer

Weekday AI; India; Remote; INR 5,000,000 - 10,000,000; Full time

Job summary

Weekday AI is looking for a senior production engineer to own the full stack of the reader product, from database to the screen. You will work directly with founders and the founding product designer to ship the product, and iterate quickly from real reader data.

You will design experiments, build reliable systems with language models at the core, ensure performance and reliability, and mentor future engineers as the team grows.

Qualifications

• Non negotiable: real production engineering depth, and the ability to build with models, context, memory and evals as a natural part of how you work.
• Preferred: several years as a full stack product engineer, experience as one of the first engineers at an early stage company, and a strong technical education.

Responsibilities

• Own the production loop: data model, streaming, background work, concurrency, payments and the credit ledger, deployment, observability.
• Build benchmarks and evals: test cases for continuity, narrative quality and failure modes to guide development.
• Manage memory, context and models: story memory across turns, model responsibilities, retrieval and cost/latency considerations.
• Set the standards the next hire inherits: ensure structure around agents to keep code maintainable as you scale.

Skills

Large language models

Backend

System Design

Education

Strong technical education such as IIT, NIT, IIIT or BITS

Tools

PostgreSQL

TypeScript

Next.js

Job description

This role is for one of Weekday’s clients

Salary range: Rs 5000000 - Rs 10000000 (ie INR 50 - 100 LPA)

Min Experience: 5+ years
Location: Remote (India)
JobType: full-time

You are the first engineer. You work directly with the founders and the founding product designer to build the product, take it to market, and improve it quickly from what real readers do. You will ship the whole product, from the database to the screen the reader sees, and you will use coding agents to do most of the typing. Your time goes on the things agents are still bad at: knowing what should be built, designing the experiment that tells us whether it works, and deciding where AI can and cannot be trusted to make the call.

What you will own.

• The production loop. Data model, streaming, background work, concurrency, payments and the credit ledger, deployment, observability. Every turn must be reconstructable after the fact, because failures in a system like this are hard to reproduce.
• Benchmarks and evals. Test cases for continuity, character behaviour, narrative quality and failure modes, developed alongside the first working loop so they tell us what to build rather than only whether we built it.
• Memory, context and models. How a story is remembered across hundreds of turns, which models do which jobs, retrieval, orchestration, cost and latency. We have starting points. We expect you to make the simple approaches fail on our own test cases before we build anything bespoke.
• The standards the next hire inherits. Enough structure around agents that speed does not turn into unmaintainable code.

A world class former CTO and two senior technical advisers are on call as a sounding board. The founders come from product and growth, have worked closely with engineers before, and know how to give you what you need and then get out of the way.

How we work

The bar is a tier A consumer product: story quality, latency, typography, state and reliability all count, and we will push on all of them. We would rather test three approaches in a week than debate them for a month. We expect you to push back when we are wrong, and we will say so plainly when we think you are. You are never stuck alone: founders are available daily, advisers are a message away, and you will have the tooling, compute and model budget to run real experiments. If something is slowing you down, we remove it.

What we are looking for

• AI first. Models, context, memory, evals and coding agents are part of how you naturally think about software, and you know where they fail.
• A real product engineer. You have shipped consumer products end to end, not worked only in notebooks or research settings.
• You have built and operated a production system with language models at its core, seen it fail with real users, understood why, and been responsible for fixing it.
• You design the experiment before you reach for custom infrastructure, and you can build an eval that measures something rather than producing a reassuring score.
• Strong on databases, APIs, concurrency, background work and production reliability. Agents can write a migration; you have to know whether it can destroy fifty thousand running stories.
• Design instinct, and the confidence to make the call when nobody hands you a spec.– Genuine interest in why people come back to a product and what makes something feel good.
• You can explain a hard technical decision in plain English, and you are comfortable challenging founders and advisers.
• You want unusually broad ownership and a path to Head of Engineering as the team grows.

Requirements

Non negotiable: real production engineering depth, and the ability to build with models, context, memory and evals as a natural part of how you work.

Preferred: several years as a full stack product engineer, experience as one of the first engineers at an early stage company, and a strong technical education such as IIT, NIT, IIIT or BITS.

Stack

Web first. We currently expect TypeScript, Next.js and Postgres, frontier models over API, and background jobs for anything that should not sit on the live reader path. None of it is sacred. You will have real say in what we use, what we buy, what we borrow from existing AI systems, and what is worth building ourselves. If you would choose differently, tell us why.

Must-have skills

Large language models, Backend, System Design

Good-to-have skills

ai agents, LLM Evaluation, Full-Stack Development