Backend platforms and APIs
Fastify, Express, and Python services with clear service boundaries, authentication and authorization, audit trails, and recovery paths designed before launch.
Remote contracts · outsourced backend work · Kathmandu, Nepal
I am Sabudh Thapa, a backend and distributed systems engineer based in Kathmandu, Nepal. I take remote contracts with teams that need event-driven workers, PostgreSQL performance work, data infrastructure, and production recovery done properly, not just shipped.
What I take on
Fastify, Express, and Python services with clear service boundaries, authentication and authorization, audit trails, and recovery paths designed before launch.
Kafka workers that survive duplicate messages, retries, replay, and ordered shutdown: idempotent producers, deduplication keys, dead-letter queues, backoff, and consumer-lag visibility.
Query design, indexing, read-model separation, and routing analytical workloads to DuckDB over Parquet when row-store plans stop paying off. Screener queries went from 8-22 seconds to under 600ms this way.
Point-in-time recovery, migration safety gates for destructive SQL, OpenTelemetry tracing across service boundaries, and dashboards that prove the system is healthy rather than assume it.
How engagements work
Project contracts. A defined system or migration with a written scope, invariants, and acceptance checks. Best for a data pipeline, a worker fleet, a query path, or a recovery tool.
Part-time retainers. A fixed weekly block for a team that already has a backend and needs it to stop breaking under load, retries, or schema change.
Full-time remote contracts. Embedded in your team, your repo, your on-call, from Nepal. This is how I currently work with a US financial-data company.
Every engagement starts with a short discovery pass: read the code, list the failure modes, agree on what "done" means. You get findings before you commit to a larger scope.
Why a backend engineer in Nepal
Kathmandu is UTC+5:45. My afternoon covers the full European morning and the start of the US East Coast day, and I keep a flexible window for standups and incident calls.
Outsourcing to Nepal costs less than hiring in the US or Europe. The reason to hire me is the work itself: idempotent, resumable, observable systems with the failure path designed first.
Read the accomplishments, run the backpressure playground, or read the engineering notes before you email.
Stack
Frequently asked
Yes. Most of my contract work is remote for companies in the United States and Europe, embedded in their repositories, review process, and on-call.
Backend services, Kafka or queue-based workers, PostgreSQL performance, data pipelines and reconciliation, production recovery, and observability. Anything where correctness under retries and partial failure is the hard part.
Frontend-only work, mobile apps, WordPress or site-builder projects, and fixed-price work without access to the codebase. I will say so early rather than take the contract.
Email a short description of the system and what is going wrong. I do a discovery pass on the code, send you written findings and a scope, and work starts once the scope and rate are agreed.
Nepal Time, UTC+5:45. I overlap the full European workday morning and the first hours of the US East Coast day, and I shift hours for standups and incidents.
Yes. I have worked as the backend engineer inside multi-vendor teams, owning service boundaries, migrations, and reliability while others owned product and frontend.
A few lines about the system, the stack, and the failure you are seeing is enough to start.
[email protected]