Hi, I'm Aarsh 👋
Fullstack Engineer (M.S. Software Engineering — expected Dec 2026) with 4+ years of professional experience.

About

Fullstack Engineer with 4+ years building backend, cloud, and AI-driven systems. I ship production services end-to-end — designing APIs, hardening infra, and deploying ML-enabled features. Multiple hackathon podiums (including a track win); experienced in Python, Django, distributed services, observability, and retrieval‑augmented ML workflows.

Work Experience

Avive Solutions Inc.

Apr 2023 - Dec 2024
Full Stack Engineer
  • ⦿ Led backend architecture improvements for geospatial services by migrating core modules to PostgreSQL + PostGIS, improving map feature accuracy by 60%.
  • ⦿ Re-architected Docker pipelines using multi-stage builds, reducing image size by >90% and cutting deployment time by 40%.
  • ⦿ Deployed and operated containerized microservices on AWS EC2 with CloudWatch monitoring, achieving 99.9% uptime.
  • ⦿ Designed centralized logging and observability using the ELK stack, enabling fast search and analysis across 10M+ logs.
  • ⦿ Resolved N+1 queries and eliminated duplicate database calls, cutting slow queries and API latency by over 50%.

Zartek Technologies Pvt. Ltd.

Apr 2022 - Apr 2023
Backend Engineer
  • ⦿ Built and deployed microservice-based REST APIs powering bookkeeping and subscription workflows in production.
  • ⦿ Integrated Veryfi, QuickBooks, and RevenueCat into production backend systems, connecting document extraction, accounting, and subscription data flows.
  • ⦿ Architected authentication and push-notification workflows with Firebase Authentication and FCM across user-facing events.
  • ⦿ Implemented serverless workflows with AWS Lambda and S3 to offload processing and reduce operational overhead.
  • ⦿ Introduced Redis caching on high-traffic data paths, reducing API latency by 35%.

Fero

May 2021 - Mar 2022
Backend Engineer (Python)
  • ⦿ Engineered a logistics automation platform on Django, supporting real-time operations with RBAC and asynchronous APIs.
  • ⦿ Modeled PostgreSQL and PostGIS data layers for routing and trip planning, reducing computation time by 30%.
  • ⦿ Shipped real-time data pipelines using WebSockets and async endpoints for live delivery tracking.
  • ⦿ Connected Google Maps API and OpenStreetMap for large-scale geospatial processing.

Education

San Jose State University (SJSU)

Dec 2026
M.S., Software Engineering (expected Dec 2026)· San Jose, CA

    Dharmsinh Desai University

    Bachelor's of Technology in Computer Engineering· Nadiad, Gujarat, India

      Languages

      Python
      JavaScript
      TypeScript
      SQL
      Rust
      Java
      C++
      Bash

      Frameworks & Libraries

      Django
      Django REST Framework
      Flask
      Angular
      Next.js
      React
      FastAPI
      pytest
      Celery

      AI Engineering

      AI Agents
      RAG
      LLM Applications
      Prompt Engineering
      Vector Search
      Embeddings
      Semantic Search
      LangChain
      OpenAI API
      Claude
      Codex
      Cursor
      Model Orchestration
      Document Ingestion
      Retrieval Pipelines

      Cloud Platforms & DevOps

      AWS
      Google Cloud
      Azure
      Docker
      Ansible
      GitHub Actions

      Databases

      PostgreSQL
      PostGIS
      MySQL
      MongoDB
      GraphQL
      Firebase
      Supabase
      Redis
      Elasticsearch

      Tools

      Git
      Linux
      Swagger
      OpenAPI
      Prometheus
      Grafana
      Logstash
      Kibana
      Django Silk
      Django Debug Toolbar
      My Projects

      Check out my latest work

      I've worked on a variety of projects, from simple websites to complex web applications. Here are a few of my favorites.

      SMOKY

      SMOKY

      Autonomous DevSecOps agent that predicts production failures before code merges. Simulates future PR merge sequences, spins up isolated Daytona sandboxes, runs Opsera security and architecture scans against the future codebase, and auto-approves or blocks PRs on GitHub, all before a single line touches production.

      Python
      FastAPI
      React
      Opsera
      Daytona
      NEAR AI
      InsForge
      GitHub API
      Playwright
      DevSecOps
      FirstCommit

      FirstCommit

      AI developer onboarding agent that helps new engineers understand repositories, documentation, setup steps, and code ownership faster. Built around natural-language repo questions, retrieval-grounded answers, and guided onboarding workflows to reduce repeated senior-engineer interruptions.

      AI
      Developer Tools
      RAG
      Agents
      Documentation
      Repos
      SeeCure

      SeeCure

      AI-powered autonomous security platform combining YOLOv8 computer vision, robot patrol simulation, cloud deployment, and real-time event alerting for intrusion, fire, and altercation detection.

      Computer Vision
      YOLOv8
      AWS
      FastAPI
      Real-Time Systems
      Chat Craft (Chatbot As a Service)

      Chat Craft (Chatbot As a Service)

      Enterprise RAG platform for ingesting internal documents and delivering context-aware LLM responses. Designed retrieval, prompt construction, inference orchestration, and API workflows with room for future model and vector store upgrades.

      RAG
      LLMs
      Vector Search
      Semantic Search
      API Design
      Hackathons

      I love building things

      I use hackathons to ship ambitious ideas fast — agentic AI systems, contagion-based virality models, legal QA engines, DevSecOps agents, and IoT safety infrastructure. 5 hackathons. Multiple podium finishes.

      • I
        2nd Place

        InsForge Hackathon

        Entrepreneurs First, San Francisco, CA

        Built Viralyst, a virality simulation platform that predicts whether a founder's demo video will spread before it's posted. Extracted Video DNA (hook timing, pacing, credibility signals) via batched vision analysis and Whisper transcription, then seeded it to a population of OCEAN-matrix AI personas running a contagion cascade that outputs an R₀ score telling founders whether their video spreads or dies. Backend built entirely on InsForge: Postgres, model gateway, and vector search powering the recommendation engine.
      • A
        2nd Place

        Applied Intelligence Hackathon

        Frontier Tower, San Francisco, CA

        Built SMOKY, an autonomous DevSecOps agent that simulates future PR merge sequences, spins up isolated Daytona sandboxes, uses Opsera to predict security vulnerabilities and architectural failures, and tracks every PR decision, risk score, and sandbox run through InsForge, all before a single line reaches production.
      • C

        Cloudathon #2 by AWS, Splunk, Verkada

        San Jose State University, San Jose, CA

        Competed in the Cybersecurity track at AWS Cloudathon, placing top 10 in a CTF-style challenge that required real-time analysis of millions of Splunk logs to identify threats, extract flags, and demonstrate hands-on cloud security and log forensics under pressure.
      • C

        CalHacks 12.0

        Palace of Fine Arts, San Francisco, CA

        Built AgentAid, an AI disaster response network that turns multilingual voice and text into structured needs, prioritizes urgency, and auto-matches requests to nearby supplies in real time.
      • L
        Won DWT Track

        LLM x Law Hackathon #5

        Stanford University, Palo Alto, CA

        Led a 4-person team to build a GPT-4-powered legal QA engine using LangChain and vector search, achieving 90%+ accuracy on long-contract questions and ranking among the top teams in a 9-hour Stanford hackathon.
      Contact

      Get in Touch

      Want to chat? Just shoot me an email with your query and I'll respond as soon as possible