Skip to content
View shmidtelson's full-sized avatar
💻
I do so as much as i can
💻
I do so as much as i can

Organizations

@Todchuk

Block or report shmidtelson

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
shmidtelson/README.md

Roman Sapezhko — Performance Engineer

Python · high-load · RPS · distributed systems · fault tolerance

I help product teams raise RPS and harden production — hot code paths, CPU, RAM / memory leaks, queues, retry storms, and failure modes across services — so critical paths stay fast and stay up.

Hands-on: profiles & traces → business-ranked backlog → change the code → fault-tolerance + monitoring so load and failures don’t take you down again.

Site sapezhko.com
Book Free 30-min diagnostic
Email roman@sapezhko.com
LinkedIn linkedin.com/in/application-performance-consultant-technical-advisor
Based Cheyenne, Wyoming, United States · Remote
Open to High-load audits · fix & harden · Python backends under real traffic

Performance Engineer — high-load and reliability Book free performance diagnostic GitHub profile views


Find me when you search for

Performance Engineer · high-load · RPS · distributed systems · fault tolerance · circuit breaker · backpressure · Python backend performance · Django / FastAPI · CPU profiling · memory leak · Postgres / Redis / Celery · throughput · p95 · observability · production reliability


Outcomes (selected)

Metric Before → After Context
API load — → 500+ RPS High-load B2B platform
API uptime under load → 99.9% Reliability bar held
Throughput baseline → +35% Same platform
Message delivery at risk → 99.9% 24/7 IoT · 2,500+/min
Origin TTFB ~2.6s → ~350ms Storefront delivery / caching

Landing with offer, cases, and sample diagnostic: sapezhko.com · source: performance-engineer


How I work

  1. Find what’s slow — or fragile — CPU, RAM, RPS ceilings, queues, retries (profiles & production signals)
  2. Prioritize by business impact — conversion, reliability, infra cost
  3. Fix and harden — change code/runtime + fault-tolerance guardrails so regressions and cascades don’t return

Stack

Backend: Python, Django, FastAPI, Postgres, Redis, Celery, asyncio
Systems: load paths, queues, backpressure, circuit breakers, delivery guarantees
Runtime: profiling, memory analysis, load testing, observability
Delivery: nginx, Docker, CI/CD, performance budgets


Featured


Currently

Building in public around high-load performance and reliability for SaaS and e-commerce.
If APIs or distributed paths hit RPS ceilings or cascade under failure — book a free 30-minute diagnostic.

Pinned Loading

  1. Update refresh token Update refresh token
    1
    createAxiosResponseInterceptor() {
    2
        const interceptor = axios.interceptors.response.use(
    3
            response => response,
    4
            error => {
    5
                // Reject promise if usual error
  2. NGiNX Configuration for Vue-Router i... NGiNX Configuration for Vue-Router in HTML5 Mode
    1
    server {
    2
      listen 80 default_server;
    3
      listen [::]:80 default_server;
    4
    
                  
    5
      root /your/root/path;
  3. Leaflet.draw Russian language Leaflet.draw Russian language
    1
    // Version 1.0+
    2
    export const leaflet_russian = () => {
    3
        L.drawLocal = {
    4
            draw: {
    5
                toolbar: {
  4. csgo_join_player_telegram csgo_join_player_telegram Public

    SourcePawn 1

  5. hh-auto-bumper hh-auto-bumper Public

    Python 1

  6. jobs_tut_by_clicker jobs_tut_by_clicker Public archive

    https://jobs.tut.by/applicant/resumes clicker extension for chrome

    JavaScript 1