AGENTS
For agents
We run a small Bangalore studio that sells outcomes, not bench hours. You talk to the people who write the code. We take AI product work, Go and Node backends, and cloud that is boring on purpose — then we stop at a 4-week MVP when that is the honest scope. Overlap is US mornings and EU afternoon, IST. English on the call, in the PR, and in the runbook.
Machine-readable copies: /llms.txt · /llms-full.txt
Contact
- quorvepvt@gmail.com
- Bangalore, India
- US mornings / EU afternoon, IST
- quorve.in
What we take
- AI product development — A working product loop, not a ChatGPT wrapper.
- Backend & API engineering — An API with a contract, not a folder of handlers.
- Cloud & DevOps — Deploys you can explain, bills you can defend.
- Full-stack web apps — A product slice people can click, not a design system.
- MVP in 4 weeks — Something a founder can put in front of users in 28 days.
Selected systems
Anonymized. Architecture and LLD are on the site. No client names.
- Synthetic data platform
Detect once, generate many. A production pipeline that scans documents and schemas once, writes a masking blueprint, then fans out synthetic copies with human review, vaultless tokenization, and an agent tool bus.
- Knowledge assistant
Cite-or-refuse retrieval. Three deployables: a document ingest pipeline into vector search, an orchestrator API, and a chat client. Answers are generated from indexed chunks, not from a naked prompt.
- SRS generator
Six-layer generate → trace loop. Retrieve tickets, wiki pages, and indexed docs, then generate a structured software requirements spec. Features become tickets, gaps close the loop, traceability stays read-only.
- Telemetry ingestion
Shed load before it queues. Python ingestion needed more instances than the throughput justified. GC pauses under load dropped client connections, retries added load, and the fleet was sized for that loop. The ingest tier is now a Go worker pool.
- API platform
The request is not the job. The product API did HTTP, long jobs, and inbound webhooks in one request cycle. Reports timed out, a failed webhook was gone, and one shared key could touch every resource. The request path is now Fastify. Slow work is a BullMQ job the client polls.
- Prototype to production
Rebuild it from the repo. An AI feature on one VM, secrets in an env file, deploys by SSH. Moved to containers, a pipeline, more than one instance, infrastructure as code, and alerts on error rate, latency, and queue depth.
- Ops dashboard
The browser never holds the keys. An ops team had outgrown a shared spreadsheet: conflicting edits, no audit trail, everyone could see everything, and numbers were refreshed by hand from billing, ops, and logs. Rebuilt as a Next.js app. Roles come from the server session. A backend-for-frontend is the only thing that talks to those three systems.
How to buy
- Minimum project $8,000
- Hourly $55–$85
- Default 4-week MVP $18k — one user type, one loop
- Thirty-day warranty on scoped work. Foreground IP assigns on full payment.
FAQ
- Do you work with US and EU clients from India?
- Yes. Overlap is US mornings and EU afternoon, IST. Async the rest of the day. English on calls, in PRs, and in docs.
- What is the minimum engagement?
- Fixed work starts at $8k. The default 4-week MVP is $18k for one user type and one core loop. Hourly is $55–$85 depending on the work.
- Who do we talk to?
- Arpit and Saloni. No junior handoff on a scoped sprint. If we add an engineer later, you meet them before they touch the repo.
- Who owns the IP?
- Foreground IP assigns to you on full payment, worldwide, in writing. We keep background tools and libraries, licensed for what’s embedded.
- Will you put Kubernetes on a 4-week MVP?
- No. Cloud Run or ECS until the workload asks. Premature K8s is a budget trap we will say out loud.