Headline

Software Engineer | AI & Backend Systems | TypeScript • Go • AWS

About

Software Engineer at PULSE | Ex-Hood I'm a software engineer passionate about building scalable systems that solve real problems. Currently at PULSE, where I architect enterprise integration platforms and AI-powered solutions that handle millions of records daily for clients like Browserstack and Chargebee. 🔧 What I Do: - Build enterprise integration ecosystems connecting platforms like Jira, Linear, and Azure DevOps with 99.8% sync accuracy - Develop AI-powered multiagent systems that automate workflows and reduce manual effort by 90% - Design real-time alerting infrastructure that cuts incident detection time from 30 minutes to under 2 minutes - Implement security frameworks (RBAC) and optimize costs by 50% through smart architectural decisions 💡 Recent Impact: At PULSE, I led the development of an intelligent copilot that generates reports in multiple formats (PPT, Excel, PDF) with automated scheduling, saving 40+ engineering hours weekly. Previously at Hood, I engineered notification systems that increased user engagement by 20% and built Bloom filters that reduced AWS costs from ₹10L to ₹5L monthly. 🛠️ Tech Stack: TypeScript • Python • Java • Node.js • ReactJS • FastAPI • PostgreSQL • MongoDB • Redis • Elasticsearch • AWS • Docker • Kubernetes • LLMs 🎓 Education: Master of Science, IIIT Bangalore 🏆 Recognition: - Performer of the Quarter at PULSE (Oct 2025) - Open-source contributor (ShameBus) - Microsoft Certified: Azure Fundamentals 💻 Side Projects: Built CodeBuddy AI - an AI-powered CLI tool for pre-commit code reviews, now available on PyPI. It helps solo developers and small teams catch issues before they become PRs. Always excited to discuss systems design, AI integration, or DevOps optimization. Let's connect!

Location

Bengaluru

AI stack

anthropic apis

claude code

langchain

mcp

openai

LiteLLM

Work experience

Senior Software Engineer at Storefox.ai, Mar 2026 - Present · 8 mo, India, 1- RAG-powered Ops Copilot: Built a RAG system over AI-generated store-conversation analyses that lets ops teams answer ad-hoc questions, explain metric drops and pull timestamped evidence clips in a single query, removing the need for new tag deployments and saving 100 hrs/week of manual clip review. 2- Multi-stage LLM Audio Pipeline: Developed a multi-tenant LLM pipeline that analyses in-store audio directly for segmentation, configurable tagging and classification, raising accuracy from 65% to 80%. 3- Scalable Audio Infrastructure: Designed containerised microservices on AWS ECS, provisioned with Terraform across multiple regions, with async workers for audio ingestion, transcription, internal/external audio classification and edge-to-cloud sync, processing 500GB+ of audio daily and reducing failed jobs by 40%. 4- Real-time Ops Dashboard: Shipped a real-time dashboard giving store operations teams a single view of AI insights, workflows, trends and device health, replacing manual data downloads and offline analysis and saving 10 hrs/week per tenant.

SDE at Pulse, Mar 2025 - Feb 2026 · 11 mo, Hyderabad, Telangana, India, 1- AI-Powered Multiagent System: Led end-to-end development of an intelligent multi-agent-powered copilot that automated query resolution, data analysis, and dynamic report generation in multiple formats (PPT, Excel, PDF) with automated scheduling capabilities for product teams, reducing report generation time by 90% (from 2 hours to 5 minutes) and saving 40+ engineering hours weekly. 2- Enterprise Integration Platform : Owned and architected 15+ upstream integrations (Slack webhooks, polling-based APIs) and downstream 2-way sync systems with Jira, Linear, and Azure DevOps for enterprise clients (Browserstack, Chargebee), handling 1M+ records daily with 99.8% sync accuracy and eliminating 85% of manual data reconciliation efforts. 3- Real-time Alerting System: Designed and built a centralised real-time alerting infrastructure across all microservices, reducing incident detection time from 30 minutes to under 2 minutes and improving system response time by 93%. 4- Role-Based Access Control (RBAC): Developed and implemented a role-based access control (RBAC) and tenant-scoped permission layer, enhancing security and ensuring seamless access management across all microservices for 50+ team members.

Software Engineer at Independent Projects, Oct 2024 - Feb 2025 · 5 mos, I’ve been actively building and launching products as a solo or lead developer, focused on solving real-world problems for both developers and everyday users: 1- DiscountBuddy – A platform that enabled users to access credit card discounts without owning the actual card. The goal was to unlock unused benefits for a broader audience. Discontinued due to tight margins and regulatory hurdles. https://www.discountbuddy.in 2- StockWiz – A comprehensive stock research platform designed to consolidate all critical stock information in one place, helping users make faster, more informed investment decisions. The project was halted due to SEBI compliance requirements and the lack of an RIA license. https://github.com/UP11SRE/stock_scraping_agent.git 3- CodeBuddy AI – An AI-powered code review tool that helps developers catch bugs, optimize code, and follow best practices before committing to GitHub. Built especially for solo developers and small teams seeking fast, actionable code reviews without relying on traditional PR workflows. This tool is currently active and evolving. https://pypi.org/project/codebuddy-ai/

Software Engineer at Hood, Dec 2022 - Sep 2024 · 1 y 9 mo, Gurugram, Haryana, India, 1- NotificationClubbing: Engineered and implemented Notification Clubbing Architecture in the backend for Broadcast and Activity-based notifications, yielding a 20% boost in engagement. 2- Fav Filter: Innovated a Deletion-enabled Customized Bloom Filter, cutting AWS costs from 10 Lakhs/month to 5 Lakhs/month by efficiently managing user actions and block IDs in Redis storage. 3- Search Service: Led the creation of the Search Service microservice from inception, seamlessly integrating Elasticsearch to optimize search functionality for existing and prospective users. 4- Redis Migration: Write multiple JavaScript scripts for efficient data migration from Redis to MySQL, realizing a notable 50% cost reduction in the Redis cluster. 5- Lambda Function: Implemented AWS Lambda for OTP verification, ensuring seamless authentication and managing traffic spikes in JavaScript. 6- Admin Panel: Revitalized admin panel with React.js, Express.js, and Mysql-Elasticsearch integration. Enhanced usability for moderators, yielding a 30% increase in Daily Active Users (DAU)).

Education

Master of Science - MS, Computer Science, International Institute of Information Technology Bangalore

Bsc hons, Information Technology, Kurukshetra university

Public links

github: https://github.com/nkansal1101