Web applications
Dashboards, admin tools, customer portals and multi-tenant SaaS. Tested at the boundaries, and built so the next team can move fast on it.
Engineering · Selective engagements
One outside problem at a time.
Most of what we build is our own. We still take on a small number of outside engineering problems each year, when the problem is genuinely difficult and we are the right people for it. This page is how that works, and how to tell if yours is one of them.
15+Years senior experience
50+Systems in production
6Industries worked in
1Engagement at a time
What we take on
Good problems, and the ones we pass on.
We are not trying to be available for everything. The work we are best at has real stakes: money moving, sensitive records, something that has to keep running while it is being fixed.
A good fit
- Systems that hold or move money. Ledgers, payments, OTC desks, custody, escrow.
- Software that keeps sensitive records, where access control and audit trails are the point, not an afterthought.
- Backends that have to keep running for years: queues, pipelines, integrations with other people’s systems.
- A platform that has outgrown how it was first built, and needs understanding before it needs changing.
- A first version that has to survive real users, not just a demo.
Not a fit
- Staff augmentation, or engineers by the hour.
- Unpaid prototypes for unfunded ideas.
- Anything built to trick people.
- Several projects at once. We take one.
If it is not a fit, we will say so on the first call, and happily point you to someone better.
Things we know how to build
Fifteen years, in production.
Built in banks, exchanges, hospitals, logistics and SaaS, and most of it still running. Client names stay under NDA; we are happy to arrange a reference call.
Custodial crypto platforms
OTC desks, per-user deposit addresses, automated hot-to-cold sweeps, programmatic escrow with dispute resolution. Custody, hardened for production.
Enterprise portals
Single sign-on, role-based access and audit trails, built so legal does not lose sleep and the ops team actually uses it. Every door closed unless someone is meant to be through it.
APIs and backend systems
Event-driven services, durable queues, real-time pipelines. Boring on purpose and bulletproof in practice — the kind of thing that sits there humming for years.
Cloud and deployment
Infrastructure as code, observability and cost discipline. Your platform in the cloud without the surprise October bill, with runbooks written before launch.
// We work with modern, proven languages, frameworks and cloud platforms, and with AI every day. The stack is chosen after the problem is understood. It is a decision, not a menu.
How an engagement works
Three phases. No surprises.
Discovery to launch typically runs seven to fourteen weeks. Every phase has written acceptance criteria, and you can stop at the end of the first one with the documents in hand.
Phase 011–2 weeks
Discovery
We sit with your team, audit the codebase if there is one, map the data, and write the smallest spec that will not lie to us in week six. Nothing is built before it is understood.
- Written specification
- Architecture document
- A clear go / no-go
- Fixed price, fully refundable
Phase 024–10 weeks
Build
Schemas, failure modes and deploy topology are agreed before production code exists. AI handles scaffolds, refactors and test coverage — every pull request still walks through a senior engineer’s hands before it lands.
- Milestone-based invoicing
- Staging from week one
- Architecture signed off before code
- Every PR read line by line
Phase 031–2 weeks · then optional
Handover
Staged rollouts, feature flags, dashboards lit up, on-call handed over cleanly. We do not celebrate at launch — we celebrate at week four with no incidents.
- Migrations, backups, monitoring
- Runbooks and on-call handover
- Optional monthly retainer
- Clean exit whenever you want it
Terms
Honest milestones, written down first.
Fixed price where the scope is knowable, milestones where it is not, and a retainer only if you actually want one. The SOW always wins over anything on this page.
CurrenciesINR · USD · EUR · GBP
Payment termsNet 14 days
TaxesGST · EU reverse-charge
RailsNEFT · Wire · UPI · Crypto
Discovery is fully refundable.
If you do not want to continue at the end of it — for any reason — the engagement ends and the fee is returned. You keep the documents either way.
Every milestone is acceptance-gated.
We invoice only after you sign off in writing. If we miss the bar and cannot fix it inside the agreed cure period, the engagement closes without final payment.
Terms are agreed up front.
Kill fees and refund terms are written into your SOW before anyone starts — never improvised after the fact. We have signed enough of them to know which ones are fair.
Get in touch
Tell us what is breaking.
A paragraph is enough: what the system does, what is going wrong, and what happens if it stays that way. We reply within one business day, always from the engineer. The first call is thirty minutes and costs nothing.
Write to us
info@larissainfotech.com
Where we are
302, Techno Park, Andheri East
Mumbai – 400069, Maharashtra
India · and remote, everywhere else
Mumbai – 400069, Maharashtra
India · and remote, everywhere else