A·MIST Résumé ↗

Makasare

Backend-focused full-stack engineer. I design, build and ship complete products — and the unglamorous systems underneath them: queues that never lose a job, locks that never deadlock, infrastructure that holds at 3am.

Builds
Backends, real-time systems & cloud infrastructure
Ships
End to end — architecture, CI/CD, deploy, load test
Based
India — remote-ready

Open to SDE-2 & senior backend roles

See the work Résumé ↗

The short version — from production, not side projects

10,000+
concurrent users on a real-time exam platform I architected
70%
lower infrastructure cost from a dual-mode, auto-scaling AWS design
6×
faster deploys — ~30 minutes down to under 5 with CI/CD
0
duplicate charges, with idempotent payment webhooks
Selected work · 03

  1. Fig. 01rest

    01Real-time · Multi-tenant · Infra

    QuizBuzz

    A multi-tenant platform for large, live, proctored quiz contests — payments, WebSockets, queues and auto-scaling infrastructure, built solo from the first diagram to the load test.

    7,500
    concurrent sockets, load-tested
    $37
    a month at idle
    < 100 ms
    API at peak
    Read the case study
  2. Fig. 02rest

    02SaaS · Payments · Async

    SmartFormFlow

    A multi-tenant event platform that replaced paper, spreadsheets and manual payments for a client — registrations, verified payments, certificates and reminders, run by background workers.

    −35%
    manual reconciliation
    −40%
    organiser follow-up
    0
    duplicate-charge incidents
    Read the case study
  3. Fig. 03rest

    03AI agents · GitHub App · In progress

    PR Sentinel

    An AI pull-request reviewer, running as a GitHub App. A verified webhook lands on a BullMQ queue; an orchestrator plans the review, runs specialised agents in two parallel waves with 12 permission-scoped tools, filters false positives in a reflection pass, and posts line comments back to the PR.

    11
    specialised agents, in parallel waves
    12×
    fewer GitHub API calls per review
    7 → 68
    tests
    Read the case study
Experience

Education
MSc Computer ScienceKK Wagh Education Society, SPPU · 2024 — 2026CGPA 8.91
BSc Computer ScienceKK Wagh Education Society, SPPU · 2021 — 2024CGPA 8.0

Software Engineer

Jan 2025 — Present

YSM Infosolution — Nashik, India · Hybrid

  • 10,000+concurrent users on a real-time examination platform, across two major versions — Socket.IO with a Redis pub/sub adapter behind ALB sticky sessions.
  • −70%infrastructure cost: a ~$37/month idle mode that auto-scales into a full load-balanced cluster within 10 minutes of a live contest.
  • hours → minto investigate payment and subscription issues, with a multi-tenant SaaS and a central operations dashboard.
  • 0duplicate charges — idempotent Razorpay webhooks for every payment and payout, and BullMQ workers issuing certificates to 10,000+ participants.
  • 6×faster deploys (~30 → < 5 min) with GitHub Actions, Docker, ECR and EC2, plus CloudWatch / SNS alerting.
About
Austin Makasare, rebuilt as a voxel portrait
Fig. 5 — The author, rebuilt in voxels.

I'm a backend engineer finishing an MSc in Computer Science. I build production systems — job queues, real-time APIs, auto-scaling infrastructure — and I'd rather a system be boring and correct than clever and fragile.

Building
SmartFormFlow — adding multi-step conditional logic and a real-time split-pane form preview
Reading
Distributed systems design — reading Designing Data-Intensive Applications
  1. 01

    Reliability over features.

    A feature that fails silently is worse than one that never shipped.

  2. 02

    Design for failure.

    Networks partition, disks fill, processes die — every system assumes the worst path.

  3. 03

    Idempotency is non-negotiable.

    An operation is safe to run twice, or it does not ship.

  4. 04

    Observability by default.

    Logs, metrics and traces ship with the feature, not after the incident.

  5. 05

    Simple scales.

    The most scalable system is the one you can still reason about at 3am.

Capabilities

Languages
TypeScriptQB · SFF · PRS
JavaScript
SQL
Backend
Node.jsQB · SFF · PRS
ExpressSFF · PRS
REST APIs
Socket.IOQB
Zod
Data
PostgreSQLQB · SFF · PRS
PrismaSFF · PRS
RedisQB · PRS
Async & distributed
BullMQQB · SFF · PRS
Redis pub/sub
Event-driven design
Idempotency
Cloud & DevOps
AWS — EC2, ALB, ASG, ElastiCacheQB
TerraformQB
DockerSFF
GitHub Actions
CloudWatch
Practices
Load testing (k6)
Infrastructure as code
CI/CD on every merge
Alerting before users notice
System design
Frontend
Next.jsSFF · PRS
React
Tailwind CSS
Framer Motion
Contact

Press the key. Every request gets a receipt — and a reply within a day.

Or directly: austinmakasare00@gmail.com

AM-200 · thermal