Senior Platform Engineer, Full-Stack

RISEUP@work

India, remote. 12:00 to 20:00 IST, because our engineering leadership sits in Portugal. Rs 2,00,000 per month plus GST if applicable. A three-month independent contractor engagement at full-time hours, not employment. Reports to Dennis Zebregs, CTO. You will need reliable high-speed internet and a setup that holds up on video calls.

The problem we are solving

Most professionals do not build a career. They react to one. They take the job that appeared, wait for the promotion that was implied, and learn the unwritten rules by breaking them. RISEUP@work closes that gap. Not job search, and not executive coaching, which arrives too late for the people who need it most. The decade in between, where a career is made or quietly lost.

A career operating system: an AI diagnostic engine, a course player, an interview simulator and an AI career coach.

You are probably somewhere inside this problem yourself. Most engineers are. Some years in, doing good work, quietly unsure whether the thing that got you here gets you to the next place. If that landed, you understand the product better than most people we will interview.

Where we are

We just deployed. What you see today is the beginning of this product, not the end of it. Real subscriptions run through this code, which is why billing and audit correctness carry the same weight as any feature. Founded by Dr.

Deepak Bhootra: thirty years of executive leadership across four countries, Fortune 100, ICF-certified coach, roughly 1,500 sessions behind the method. Over a thousand professionals have signed up. We are funding through a community round at wefunder.com/riseupatwork. The roadmap is long and mostly unbuilt, and none of it ships without engineers.

The role

You will work across our API service and web app, and own features end to end. Schema change, endpoint, screen, deploy. Backend one week, React the next.

You would report to Dennis Zebregs, our CTO and the lead developer on this codebase today. Your judgment carries weight from week one, and nobody hands you a queue of tickets. The stack below is broad, so apply even if you see yourself in most of it rather than all of it. Infrastructure ownership stays with the CTO for now, so you ship through an existing Docker and GitHub Actions pipeline rather than authoring Terraform.

What you will do

- Ship backend features in NestJS: controllers, services and modules, plus Prisma schema changes and migrations against PostgreSQL, including our pgvector embedding tables.
- Wire modules together through a typed application event bus rather than direct cross-module imports.
- Build LLM-backed features: the RISE coach agent, AI diagnostic report generation, interview simulator feedback. These route through our internal model router, not a provider SDK directly.
- Build member-facing UI in Next.js App Router and React: dashboard, diagnostics, Knowledge Tracks, interview simulator, wired to the API through a single authenticated proxy layer.
- Carry a feature through its lifecycle: migration, endpoint, screen, image, staging, production.
- Debug production issues end to end, including direct database access under a documented, access-controlled procedure.
- Keep our contributor docs current. A gotcha that has gone stale, or a current one you hit, gets written down.
- Raise what you find. Bug, security gap, stale doc, infrastructure risk, design smell:



file a GitHub issue with the existing priority labels. Staying quiet about a known problem is worse than an issue that turns out to be nothing.

Our stack

Backend: TypeScript, Node.js, NestJS, PostgreSQL with Prisma ORM including vector and embedding search, event-driven module architecture, LLM integration, Redis, Jest.

Frontend: TypeScript, Next.js App Router, React, Tailwind CSS, Cognito-based auth.

Infrastructure, owned by the CTO today: Terraform and OpenTofu, AWS (ECS, RDS, Cognito, Secrets Manager, load balancing), Docker, GitHub Actions.

We run recent major versions, so expect some APIs and conventions not to match older docs.

What you will need

- Five or more years building production software with Node.js and TypeScript in strict mode, including a dependency-injection-based backend framework. NestJS, or something close enough to transfer.
- An ORM and migration workflow against PostgreSQL, Prisma or equivalent, including reasoning about indexes and query plans.
- Production React with a modern meta-framework. Next.js App Router strongly preferred.
- Hands-on experience putting an LLM into a shipped feature: prompt and response handling, retries, cost and rate-limit awareness.
- JWT-based auth in production, issuing, verifying and refreshing, ideally against a managed provider such as Cognito or Auth0.
- Rigor around money and audit trails. Usage metering, subscription state, payment webhooks. Recording an effect only after the operation it represents has succeeded.

Bonus points

- Event-driven module boundaries inside a monolith. Vector and embedding stores, LangChain or LangGraph.
- Data-fetching and UI state without a state or query library by default. We write caching, retry and de-dupe by hand where that is right.
- Utility-first or token-based CSS and the discipline to use design tokens rather than one-off values.

Browser auth flows: OAuth and OIDC, httpOnly session cookies, silent token refresh.

- Docker builds and shipping through an existing CI/CD pipeline. Terraform or OpenTofu, plus core AWS primitives (ECS, RDS, IAM, Secrets Manager). A plus, never a gate.

How we work

- We build small and deliberately. Fewer dependencies, more hand-written code in places. A hand-rolled markdown renderer, a hand-rolled WebRTC hook. A decision, not an oversight. Know when thirty lines of plain code beats a package.
- Our own docs are the source of truth, above general best practice and above what worked at your last job. When they are wrong you fix them, and when you hit a deploy trap you write it down.
- Tooling maturity varies. Some areas lean on a documented pre-commit checklist rather than full CI-gated lint and test coverage, so you hold that bar yourself. Expect scoped production database access under procedure, and infrastructure state that does not always match git.
- AI-assisted, not AI-driven. Built with Claude Code and per-repo conventions docs. Review AI-generated changes as you would a colleague’s pull request, and never let an agent make an architecture,



security or scope call that is yours.
- Someone’s career runs through this. The person on the other end is making real decisions about their working life using what we build. That raises the correctness bar past what an internal dashboard demands.

The engagement, and what it pays

What we commit to: three months as an independent contractor at full-time hours, Rs 2,00,000 per month plus GST if applicable, paid monthly against your invoice and fixed. If you are weighing this against a salaried offer, run the numbers rather than the headline. Professional fees are taxed very differently from salary in India, so the take-home lands closer to a materially larger CTC than the monthly figure suggests.

What we are not committing to: employment. Not on day one and not automatically at the end. The honest reason is that our Indian entity is registered but the payroll strategy around it is not finished, and we will not promise what we cannot deliver. For these three months you invoice us, handle your own tax and keep your own books. No PF, no gratuity, no notice period. Satisfy yourself on withholding, advance tax and GST before you sign.

After three months we will look at employment once the entity strategy is locked, and at a longer contract, which for an India-based engineer is often the better of the two once tax is accounted for. What we will not do is string you along on rolling three-month contracts. The decision sits with Dennis and it is a judgment call, not a scorecard: technical competence on this codebase, seeing past the ticket to what the product needs, and how you work in a team this small.

You will know where you stand well before it ends.

On equity, for the right person we would consider a stake in anticipation of a future event. We cannot commit to it and will not put a number on it here, so do not take this engagement because of it. If you cannot start on contract, this is not the right role for you.

Before you apply, go and use it

Sign up at riseupatwork.com. Free, and every sign-up gets a 14-day Builder trial, no card required. That is the full product. Run the diagnostic on yourself.

Do not look at where we are. Think about where you can take us. You will find rough edges and we would rather you found them now. But the applicants who make an impression are not the ones who send us a defect list. They used the product, understood what it is reaching for, and came back with a point of view.

How to apply

Keep it short. Two things matter.

Show us something real. Shipped work, a repository, a pull request, a technical writeup, a talk, a public profile. Anything verifiable beats a list of adjectives. If the work was private, describe the system and your part in it well enough that we can ask a real follow-up question.

Tell us what you would build. One short paragraph. Not a critique of what exists. A direction, if the roadmap were yours.

Send it to [email protected] and [email protected], and tell us which side of the stack you are strongest on.

Interview process

- Thirty-minute intro call with Dennis.
- A technical conversation on something you have built end to end.
- A paid assignment. Everyone at this stage gets the same brief and seven calendar days. We may run several of you through it at once, and will tell you if so.
- Final conversation on scope, growth and offer.

Expect around three weeks from first call to decision.

📌 Senior Platform Engineer, Full-Stack (India)
🏢 RISEUP@work
📍 India