Open to new opportunities

Backend & distributed systems engineer.

I'm Paritosh Tripathi — a backend / distributed-systems engineer. Most recently Software Engineer II at Parspec, where I built high-throughput, event-driven backend services — REST APIs, ERP integrations, distributed pipelines, and platform infra used across engineering. Previously Software Engineer I at Anakin (YC S21). Currently open to new opportunities.

Now › Most recently Software Engineer II at Parspec (through Jun 2026) — now looking for senior backend / platform / distributed-systems roles.

paritosh@portfolio · ~ type 'help'

~ $

See selected work → Resume ↗ GitHub ↗

— Selected work

Five systems I designed end-to-end

Click in for architecture, decisions, and the numbers.

Iris

Per-branch ephemeral environments. PR open to live URL in 3.8 min, host-header routed on a shared ALB.

1,567: builds
98.5%: build success
88.1%: end-to-end
7×: DAU growth

FastAPI ECS Fargate ALB CodeBuild ECR Route53 Postgres

Read the case study →

Argus

Saga with compensating transactions and Redis-backed per-SKU locks behind the inventory ingestion path.

3.9M: writes
7,800: manufacturers
30%: MoM growth
407K: peak day

Django REST DynamoDB Redis Saga Per-SKU locks

Read the case study →

Ariadne

Soft-filtering search with weighted ranking, match-strength buckets, and ML image similarity. Held to per-tier p90 latency budgets.

21%: accuracy lift
1.5s: p90 (top 50)
5×: mounting weight
3: match buckets

Elasticsearch OpenSearch Vector search Python ML similarity

Read the case study →

Proteus

Config-driven FastAPI service that replaced multiple per-team custom compilers — JSONPath configs, dual-queue SQS + DLQ, multi-sink writes.

200 GB: peak day
100+: concurrent tasks
3d → 4h: client onboarding
99.95%: availability

Python FastAPI SQS ECS MongoDB BigQuery S3

Read the case study →

Hydra

Hybrid Pixel + Genymotion automation platform with affinity scheduling and runtime Frida injection.

10: concurrent sessions
100+: devices managed
Hybrid: on-prem + cloud
Frida: runtime hooks

Python Go ECS SQS PostgreSQL MongoDB Frida Genymotion

Read the case study →

— Timeline

Where I've been

Years mark the line; roles alternate along it — click one for the full writeup.

2026

2025

Apr 2025 — Jun 2026

Software Engineer II

Parspec

Led the redesign of a core inventory microservice — 10K+ distributed operations under 300s at 99.6% success via a SAGA-based distributed locking system.
Delivered REST APIs (Django, FastAPI) on DynamoDB / PostgreSQL / OpenSearch serving 100K+ items/hour with sub-100ms P99 latency via Redis caching.
Built an internal automation platform — 100+ isolated PR environments — cutting setup time 95% (4h → 8min) and lifting team velocity 30%.
Owned 20+ production APIs handling 50K+ daily requests at 99.8% uptime, monitored via CloudWatch and custom metrics.

2024

2023

2022

Aug 2022 — Dec 2024

Software Engineer I

Anakin (YC S21)

Architected a fault-tolerant distributed pipeline processing 150 GB/day across 4 teams (AWS SQS / S3 / ECS, MongoDB, Docker, Jenkins) at 99.95% availability.
Cut reprocessing overhead 15% and infra costs $3K/month; built Looker Studio dashboards on BigQuery cutting defect resolution time 29%.
Designed event-driven microservice architecture — 150M+ events/month at 99.9% uptime, 50K events/min peak — via partitioning and async processing.
Engineered reusable microservice templates and DDD patterns adopted by 6 teams — feature dev time 2 days → 3 hours, scaling 12+ services independently.

AWS

Python