parcel-router01 / 06
Deterministic carrier routing
34% → 6%
Portfolio — 2026Lublin, Poland · remote Available for work
Senior Backend Engineer · Logistics & parcel SaaS
WladyslawCzeresniowski
I build logistics backends that move parcels, money, and freight — Python/Django systems, carrier integrations, Postgres-heavy workflows, and Rust where throughput matters.
I build the event-driven backends that move money — price streams, courier integrations and reconciliation, and the automatic refunds that have to be boring in production.
By the numbers — reproducible benchmarks 01 / 05
34% → 6%
mis-route rate — the carrier-routing engine on cost, SLA, and live capacity, against a hand-tuned-rules baseline, in a reproducible benchmark.
~26×
lower p99 task time after rewriting the Django/Celery queue worker in Rust — the headline of the migration write-up.
8M
freight rate-quotes a day in the Go rating service's load profile; its committed load test sustains ~46k req/s, far above that.
<40ms
freight rate-quote p99 in the Go rating service — its load test measures ~11 ms at ~46k req/s, well inside the 40 ms budget.
0
double-refunds by construction — a row lock and an idempotency key in the refund engine, shown by its runnable incident demo.
01
Open source · reproducible benchmarks
Selected projects & benchmarks
freight-rating02 / 06
Freight-quote rating service
p99 < 40ms
skip-locked03 / 06
Postgres as a job queue
one table
refund-engine04 / 06
Idempotent refund pipeline
no double refunds
idem-key05 / 06
Idempotency keys, done once
Rust + Go
pg-outbox06 / 06
Transactional outbox
same-tx
02
Experience — 10+ years building backend systems
2018 — Jul 2026
Staff Backend Engineer (B2B contract)
Pakke.dk · Python · Django · Kubernetes
- Rebuilt the quote engine that prices every shipment across 8 carriers — ~10k quotes a day, 50k at the December peak — taking response time from ~10s to ~100ms.
- Built tracking and SLA monitoring over tens of thousands of carrier events a day, with automated refund claims on every breach — gross margin up ~5–7 points.
- Moved quoting to Kubernetes with autoscaling and zero-downtime deploys, set up Prometheus, Grafana and Loki, and moved carrier integrations from XML/SOAP to REST.
- Hired and onboarded 4 engineers and ran code review across the Polish and US teams.
100×faster quotes · 10s → 100ms
8carrier integrations
4engineers hired
Carrier routing engine
2024 — 2026
Founder & Sole Engineer
remitto.io · formerly Refundify · Django → NestJS · Temporal · Postgres
- Built it solo, end to end: invoice parsing across carrier formats, line-item matching against shipments, and delivery checks against carrier tracking APIs.
- Rebuilt the claim pipeline as durable Temporal workflows — detect the breach, file the claim, reconcile the payout — idempotent at every step so a retry never double-claims.
- 15 EU customers, mostly Nordic; ~100k parcels audited a month; over $2M recovered for clients.
$2M+recovered for clients
100kparcels audited / month
15EU customers
Idempotent refund state machine
2015 — 2018
Co-founder & Engineer
Rise & Raise · Django · Node · Postgres
- Built the real-time pricing pipeline across ~40 exchange and marketplace APIs, normalized into one live stream.
- Fed the trading bots off price events — thousands to tens of thousands of trades a day, no polling loops to fall behind.
- Built the customer web app: registration, billing, and subscriptions.
40exchanges & marketplaces
400paying subscribers
1live price stream
Live price-stream aggregator
04 — Get in touch
Open to senior & principal backend roles
Get in touch.
If you are hiring for a senior or principal backend role in logistics, payments, or anything high-throughput, email me. One honest reply, no recruiter funnel.
wladyslaw@czeresniowski.devBased in Lublin, Poland · remoteReplies within a dayAvailable now