Hi, my name is

Harshit Rathore.

I'm a backend engineer specializing in building scalable microservices, distributed systems, and AI-enabled platforms. Currently, building at Techolution.

Download Resume

01. About

I'm a Software Engineer who cares about building systems that are reliable under real production load. My expertise spans backend engineering, distributed systems, microservices architecture, and cloud infrastructure on GCP.

At Techolution, I've architected and shipped production systems end-to-end: an AI-enabled code modernization platform, a fault-tolerant distributed scheduler processing 10,000+ jobs, an event-driven notification pipeline, and an agent-based text-to-SQL system used by 100+ business users. I've also debugged and fixed hard production issues — including a MongoDB connection leak that was silently doubling API latency.

I think in terms of measurable impact: latency reduced, throughput doubled, time-to-market cut by weeks.
Outside of work, I've solved 400+ problems on LeetCode (Contest Rating 1639, top 22% globally) and stay sharp on data structures, algorithms, and system design.

Harshit Rathore

Here are a few technologies I've been working with recently:

02.Where I've Worked

Backend Engineer @ Techolution

July 2024 — Present

  • Architected and led backend development for StackMod, an AI-enabled code modernization platform, cutting manual refactoring effort by 40%; built an automated validation pipeline generating and running unit tests against legacy vs. modernized code, lifting AI task accuracy by 35%.
  • Redesigned the notification delivery pipeline from DB polling to event-driven Redis Pub/Sub + SSE, cutting latency by 50% and DB load by 80% — including a dedicated SSE gateway with per-user connection queues for real-time delivery across multi-pod GCP infrastructure.
  • Designed and built a fault-tolerant distributed Scheduler Microservice adopted by 10+ product teams, powering 10,000+ scheduled jobs with exactly-once execution guaranteed via Redis NX locking.
  • Built an agent-based text-to-SQL system on Gemini Flash Lite, automating query generation, execution, and AI-backed response synthesis — cutting manual reporting effort by 80% for a 100+ person business team.
  • Engineered Role-Based Access Control (RBAC) with audit trail logging for a clinical trials platform, enabling granular access across multiple user roles and strengthening regulatory compliance.
  • Diagnosed a MongoDB connection leak causing 60+ second API hangs by tracing 800+ dangling connections — cutting CRUD API latency by 60% and doubling peak throughput.
  • Built a modular Notification Microservice with a plug-and-play architecture (Slack, Outlook, and more), accelerating time-to-market for 4+ product teams by ~1 month each.

Backend Engineer Intern @ Techolution

Nov 2023 — June 2024

  • Led a refactoring initiative that improved development speed by 80%.
  • Built "PR Reviewer" — an automated code-review tool that reduced production bugs by 50%.
  • Built RAG-based systems and contributed to core backend infrastructure improvements.

AI Intern @ AIoT Technologies

Apr 2023 — Sep 2023

  • Developed machine learning models using Python, TensorFlow, and CNN architectures.
  • Contributed to AI/IoT projects exploring deep learning and neural network optimization.

03. Some Things I've Built

01 / 03
Online
product discovery & ordering
  • Python
  • FastAPI
  • MongoDB
  • Motor
  • Cloudinary
  • HTTPX

CrackerKart

E-commerce & Inventory Platform

Problem

A traditional fireworks business faced stagnant revenue and limited online visibility, making it difficult for customers to discover products, compare pricing, and check discounts and stock.

Solution

Built the complete async backend and database layer for an e-commerce platform with centralized pricing, multi-variant inventory, reliable order processing, admin operations, and automated customer communication.

Features

  • Multi-variant catalog with centralized pricing, discounts, and real-time inventory management
  • Reliable order workflow with validation, pricing calculation, stock deduction, and status tracking
  • Automated customer notifications with transactional email and localized WhatsApp messaging

Impact

Digitized product discovery and ordering, eliminated manual pricing calculations, reduced inventory inconsistencies, and automated customer communication for seasonal operations.

02 / 03
Personalized
context-aware AI assistance
  • Python
  • Flask
  • Gemini
  • Gmail API
  • Firebase
  • Google OAuth
  • Docker

Mango AI

Cross-platform Conversational Assistant

Problem

Traditional phone assistants handled isolated commands, while users needed conversational, multi-step assistance for tasks like understanding emails, drafting responses, and learning concepts.

Solution

Built Mango AI, a cross-platform conversational assistant combining voice interaction, Gemini-powered reasoning, persistent conversation state, and Gmail intelligence through an API-driven Python backend.

Features

  • Context-aware AI personas with persistent multi-turn conversations
  • Gmail intelligence for email summarization, categorization, and AI-generated responses
  • Voice-first interaction with structured AI responses and secure Google OAuth integration

Impact

Automated repetitive email workflows and demonstrated an end-to-end conversational AI system spanning voice input, backend orchestration, LLM reasoning, state management, and structured user-facing responses.

03 / 03
Centralized
institutional contact management
  • Python
  • Flask
  • MySQL
  • SQLAlchemy
  • Pandas
  • OpenPyXL

Contactly

Collaborative Contact Management Platform

Problem

Years of student and parent records were fragmented across individual devices, making historical data difficult to access, maintain, and share reliably across faculty and administrators.

Solution

Built Contactly, a centralized contact management platform with secure multi-user access, structured data storage, group-based organization, granular sharing, and bulk Excel data ingestion.

Features

  • Centralized contact repository with rich student, parent, address, and professional metadata
  • Group-based organization with granular view/edit permissions for controlled collaboration
  • Bulk Excel ingestion and token-based authentication for efficient, secure data management

Impact

Established a single source of truth for institutional contacts, reduced manual data maintenance through bulk ingestion, and enabled secure collaboration while preserving historical records across academic years.

04. Othello AI (try it out!)

Your turn

White (You): 2 Blue (AI): 2

05. What's Next?

Get In Touch

I'm excited to explore opportunities where I can work on technically challenging products where strong engineering translates into real-world impact — from designing robust backend systems to building and shipping AI-powered experiences. Whether you have a question, a new opportunity, or just want to talk backend engineering, I'd love to hear from you.

Email: harshit.rathore.dev@gmail.com
Mobile: +91 99265 46160

Say Hello