Education and Experience
Pardy Panda Studios
Software Engineer I
InfraCorridors
AI Infrastructure Marketplace: Built the complete platform for a global AI and telecom infrastructure marketplace : covering the admin portal, client-facing UI, and all backend APIs. Enterprise users can browse verified vendors across geographical corridors, research subsea cable routes, and submit procurement RFPs through a streamlined bidding flow. Leveraged React Server Components with Promise.all parallelization to achieve sub-500ms TTFB across pages that aggregate vendors, corridors, and blogs in a single server request cycle. Designed a hybrid data architecture separating core domain data (external microservice API) from lead data (local MongoDB Atlas), keeping high-value RFP submissions isolated and secure. Implemented reCAPTCHA v3 on all submission endpoints to protect against bot-driven spam on a public B2B procurement surface.
S3 Cable Data Sync Pipeline (lives inside InfraCorridors admin)
Engineered a TypeScript sync service that fetches live submarine cable data from the global cable map API and uploads 5+ structured JSON files per sync cycle to AWS S3 : including per-cable geometry, landing points, cable index, and a merged full-dataset file. Built MD5 hash comparison via S3 ETag diffing so unchanged files are skipped entirely, reducing redundant S3 bandwidth to zero. Implemented concurrency-controlled parallel fetching with retry logic across all cable endpoints. The sync is triggered from the InfraCorridors admin panel and the map tool reads directly from S3 : no database dependency for geodata at the frontend level.
InfraCorridors Map Tool (separate public product, same client)
Built the public-facing submarine cable map tool as a standalone product. All geodata : cable geometry, landing stations, and cable index : is served directly from S3, making the tool fully serverless with zero backend calls at runtime. Users can explore global submarine cable infrastructure visually through an interactive MapLibre-powered interface. Deployed independently from the admin system, with ETag-based browser caching ensuring users only download fresh data when the underlying cable data actually changes.
Stripe Marketplace
Payments Architecture: Architected a full Stripe Connect marketplace supporting 4 service types (Blog, Q&A, Video Call, Chat) with seller sub-wallets under the platform account using Express Connected Accounts. Built a slab-based commission engine calculating platform fees across 4 tiers (3% → 10% based on order value up to $1500+), applied at transfer time using idempotency keys to prevent double payouts. Blog purchases settle instantly; Q&A, Video Call, and Chat payments are held for 24 hours then auto-released via a cron job running every 15 minutes. Integrated 10 Stripe APIs end-to-end : PaymentIntent, Transfer, Refund, Payout, Connect Express Accounts, Account Links, and 4 webhook handlers. Routed video call meeting links through the platform server to generate timestamped access logs as dispute evidence : without any third-party call tracking.
Insansa Techknowledge Pvt Ltd
Software Engineer
Full API Ownership
Delivered all backend APIs for real-time pricing management, system configuration, and user management workflows : built production-ready with no rework cycle.
Frontend Architecture Overhaul
Refactored the frontend from a client-heavy component model to Next.js server components, restructuring into a modular architecture that reduced repetitive code by 20:30% and cut future feature build time significantly.
Cloud Infrastructure
Deployed the full system on AWS EC2, integrated S3 for media storage with IAM-controlled access, and implemented structured error logging : contributing to a 20% improvement in system uptime during peak usage.
Aubergine Solutions Pvt. Ltd
Software Engineer Intern
Payments & Workforce Systems
Implemented employee work tracking, credit-based referral management, and feedback modules. Supported production deployment via Docker and Nginx reverse proxy : ensuring a scalable, maintainable service delivery pipeline.
Frontend Performance
Participated in code reviews and debugging sessions optimizing a large-scale MERN codebase, eliminating unnecessary re-renders and contributing to faster UI rendering across the platform.
Credit-Based Referral System
Built the referral and commission tracking flow end-to-end, including Admin oversight views and client-facing reward dashboards : handling multi-role state correctly across the RBAC layer.
Academic Foundation
MCA — The Maharaja Sayajirao University of Baroda
Gujarat, India · Engineering Systems & Advanced Computing