Scale & harden your backend
Performance tuning, caching, query optimization, and CI/CD so releases stop being a gamble.
SaaS · marketplaces · subscription products
Discuss this service Open to contract & consulting · Remote worldwide
// Senior Python Developer · Backend Engineer
Backend systems that scale — without breaking in production
Core stack: Python
I help product teams turn complex requirements into stable Django/Python backends: APIs, queues, payments, real-time, and AI — shipped with tests and observability.
Seville, Spain · Remote EU / US / worldwide
Trusted by SaaS, IoT, and digital-commerce teams · 10+ years in production
python3 — repl
10+
Years in production
SaaS, IoT, commerce
99.9%
API uptime
Rate limits & monitoring
500+
Requests/sec
Peak API throughput
10k+
Users served
Multi-tenant platforms
// services
What I help you ship
Clear scope, measurable outcomes, and production-ready code — not slide-deck architecture.
Real-time & IoT systems
Event-driven pipelines, device telemetry, and reliable message queues when seconds matter.
IoT · logistics · mobility · hardware teams
Discuss this serviceAI in production — safely
LangChain integrations, guardrails, and backend workflows that connect LLMs to your data.
Support automation · internal tools · data products
Discuss this service// tech_stack
Production-proven stack
Tools I use daily to keep systems fast, observable, and cheap to operate at scale.
Backend
- Python
- Django
- Flask
- Celery
- REST API
- LangChain
Frontend
- Vue.js
- TypeScript
- Nuxt
- HTML/CSS
Infrastructure / DevOps
- Docker
- Kubernetes
- AWS
- Coolify
- Portainer
Databases
- PostgreSQL
- PostGIS
- Redis
- RabbitMQ
- ClickHouse
Observability & CI
- Sentry
- New Relic
- GitLab CI/CD
- GitHub Actions
// experience
Results in production — not resume filler
Every engagement focused on metrics your business cares about: speed, uptime, cost, and risk.
GDS Group
Senior Python Developer / Backend Engineer
2023 — 2025
Remote · Los Angeles
- Scaled a Django/Postgres/Redis digital goods platform — transaction throughput +35%.
- Cut analytics generation from 4 hours to 15 minutes (Celery + query optimization).
- APIs at 99.9% uptime with rate-limiting and real-time dashboards — 500+ req/sec.
- CI/CD and Docker: deployment time −65%, test coverage 90%.
Trust Box LLC
Python Developer · IoT & Real-time Systems
2021 — 2022
Remote · Moscow
- IoT warehouse box rental: retention +20%, Orange Pi integration, 99.8% uptime.
- RabbitMQ + Celery: 2,500+ messages/min with 99.9% delivery.
- Payment system re-architecture: support tickets −60%, GDPR-compliant design.
- Device alert response time: from 15 minutes to 30 seconds.
Kvitly (SaaS Website Builder)
Python Developer · CRDT & Multi-tenant
2019 — 2021
Remote · Warsaw
- Multi-tenant backend for 10,000+ users, REST API at 99.9% uptime.
- CRDT backend for real-time collaborative editing — up to 15 concurrent editors.
- Celery/RabbitMQ: 1,500+ background ops/min, payment failures −40%.
Fast Kicksharing
Python Developer · IoT & GIS
2018 — 2020
Remote
- Scooter geo speed limits (PostGIS), event-driven order processing architecture.
- Real-time telemetry via RabbitMQ for thousands of devices simultaneously.
- Monitoring and alerts: downtime −20%.
// featured_projects
Proof of complex delivery
Selected builds where architecture decisions directly unlocked product growth or reliability.
CRDT Collaborative Editor
Conflict-free document sync with 15+ simultaneous editors without data loss.
→ Enabled real-time co-editing for a SaaS website builder.
- Python
- CRDT
- Django
- WebSockets
- PostgreSQL
AI First-Line Support
OpenAI and Claude integration via LangChain for automated first-line ticket handling.
→ Reduced manual triage load on support workflows.
- Python
- LangChain
- OpenAI
- Claude
- REST API
IoT Warehouse Platform
Box rental automation: Orange Pi, subscriptions, RabbitMQ queues, and real-time device status.
→ Retention +20% · alert response 15 min → 30 sec.
- Django
- Orange Pi
- RabbitMQ
- Celery
- IoT
// testimonials
What teams say after shipping
Generic feedback patterns from B2B SaaS, IoT, and platform engagements — names anonymized.
“Roman took our Django monolith from “works in staging” to predictable releases. API latency dropped, on-call got quieter, and we finally trust the deploy pipeline.”
“We needed real-time device telemetry without losing messages at peak load. Queues, monitoring, and clear runbooks — production stopped being a black box.”
“Clear scope, no architecture theatre. Weekly updates, tests in CI, and backend decisions explained in business terms — not slide-deck jargon.”
// contact
Let’s talk about your backend
Tell me where you’re stuck — scaling, reliability, integrations, or greenfield architecture. First call is free, no commitment.
Typical reply: within 24 hours · English & Russian
What happens next
- 1 15-min intro — goals, stack, timeline
- 2 Scope & approach — fixed price or monthly
- 3 Kickoff — repo access, milestones, weekly updates
Or reach out directly
roman@sapezhko.com · Seville, Spain · Remote worldwide