Alireza Nasiri

Senior Software Engineer | Backend · Python & Go

↓ PDF

About

Senior backend engineer with 7+ years building and running production platforms in fintech, marketplaces, and AI, working primarily in Python and Go with a deep grasp of distributed systems, concurrency, and system design. I own services end-to-end, from architecture and implementation through deployment, monitoring, and production support, and most of my work is making systems hold up under real load: I raised a KYC service's weekly availability from 70% to 99%, scaled B2B cash-in flows 60x, and cut production incidents on a review platform serving 2,000,000+ daily requests. Recently I've been shipping LLM-backed features in production, with schema-enforced output and cost and latency tracking. I write thoroughly tested code, keep CI green, and watch my own changes in production.

Experience

Senior Software Engineer

Voicedash · AI voice dictation and meeting-notes app

Mar 2026 – Present

  • Built the database-driven model-routing engine for the transcription pipeline, selecting STT and LLM-polish providers per request with per-variant timeouts, single-level fallback, and retries, cutting p99 latency by about 33% (9.5s to 6.4s) versus the legacy route.
  • Integrated several OpenAI-compatible STT providers along with a Telegram-based shadow-evaluation system that scores alternative models on copied live traffic without changing user-facing responses.
  • Made LLM output reliable in production by enforcing a JSON schema through response_format, adding a defensive parser that never shows the user a damaged response, and keeping the system prompt static so it can be cached.
  • Instrumented the backend with Sentry tracing, per-request and LLM-cost metrics, and latency distributions tagged by output language (570K+ data points), then tuned span volume to control cost.
  • Defined the production infrastructure as code with Terraform, then planned and scripted the move from a US to an EU host with a rehearsal run, preflight checks, and a cutover runbook.

Senior Software Engineer

TomanPay · Fintech: escrow and B2B payments for marketplaces like Divar

Dec 2023 – Sep 2025

  • Revamped the core KYC verification system, raising weekly service availability from 70% to 99% by refactoring legacy code and implementing robust monitoring, enabling onboarding of 1,000 new users daily.
  • Engineered and maintained the core escrow service using Django, MySQL, and Redis, handling approximately 40 requests per second and implementing the hold-and-release flow that keeps the payer's funds in escrow until they confirm the goods are as described.
  • Built B2B cash-in services, scaling daily fund inflows 60x (from 1B IRT to 60B IRT) and reducing manual operational workload by 90%.
  • Developed an exchange and order book service in Go, benchmarked in load testing at over 1,000,000 matches per second.
  • Made money-movement flows safe under concurrency with Redis locks, task locking, and deadlock recovery, and added a zero-sum check on matched exchange orders.
  • Brought the legacy escrow codebase under test by making the suite runnable, fixing broken tests, and adding coverage reporting and a formatting gate to CI.

Senior Software Engineer

Ravindex · Fintech: stock-market and crypto analytics and trading

May 2022 – Dec 2023

  • Cut response times on the financial dashboards by removing Django ORM N+1 queries, adding missing indexes, moving the heaviest calculations into SQL-level functions, and precomputing the rest off the request path.
  • Designed a real-time data update service using socket-based connections, integrating Tehran Stock Exchange (TSE) data via an event-driven architecture.
  • Built a crypto copy-trading engine integrated with Binance, KuCoin, and three other exchanges that mirrored lead traders' positions for followers and computed trader PnL, rankings, and performance history.
  • Built fixed-income and options analytics for the Tehran Stock Exchange, including a yield-to-maturity calculator, fund revenue tracking, and option chain, spread, and covered-call strategies.
  • Built an order management system MVP, connected it to the market-data backend through RabbitMQ consumers, and set up Prometheus monitoring for both backends.

Software Engineer (Review & Car Team)

Divar · Iran's largest online classifieds marketplace (like Craigslist)

Aug 2018 – May 2022

  • Owned the review pipeline and, over time, several other backend services on a review platform that grew to around 20 microservices and bots processing 2,000,000+ daily requests.
  • Cut production incidents from about five to fewer than two per week by tracing them to a root cause in Elasticsearch indexing; uptime rose from 98% to 99% over the same period.
  • Automated CI/CD pipelines using GitOps, reducing deployment time from 60 minutes to under 10 minutes.
  • Generated 600M IRT in revenue by building and scaling Karnameh (Car Inspection Service) from the ground up, reaching 200 daily inspection requests by my departure in 2022.
  • Deployed an optimized plate detection bot, fixing memory leaks and cutting processing time by 66% with parallel image processing.

Skills

Languages

  • Python
  • Go
  • Java
  • SQL

Backend & Frameworks

  • Django
  • REST & gRPC APIs
  • Microservices
  • Event-Driven Architecture
  • Kafka / RabbitMQ
  • System Design
  • Automated Testing

AI & LLM

  • LLM / AI Agents

Databases

  • PostgreSQL
  • MySQL
  • Redis

Cloud & Infrastructure

  • Kubernetes
  • Docker
  • AWS
  • Terraform
  • Ansible
  • CI/CD (GitOps)

Observability

  • Prometheus
  • Grafana
  • Sentry

Education

Sharif University of Technology

B.Sc., Computer Science

Sep 2022

Honors & Awards