Software Engineer with 2+ years turning ideas into production systems — APIs, databases, and now LLM-powered applications running on self-hosted infrastructure. Currently building enterprise platforms at Appnox Technologies.
A little about how I got here, and what I do when I'm not shipping code.
Software Engineer · India
It started with curiosity — wanting to know how the things on a screen actually worked underneath. That pull toward "how does this work?" turned into learning to code, then into a Computer Science degree focused on AI & Machine Learning, and eventually into a career building real backend systems for real companies.
Today that curiosity shows up as a habit: every new LLM provider, every new deployment pattern, every interesting open-source tool — I want to understand it well enough to use it in production, not just read about it. That's how a CRM ended up with a self-hosted model running on RunPod, and how a customer support assistant ended up speaking three channels through one RAG pipeline.
Outside of work, I'm usually reading about whatever's new in tech — new models, new frameworks, new ways to deploy things faster — or traveling when I get the chance, because a change of scenery is the best debugger for a tired brain.
Backend fundamentals paired with hands-on AI integration and the infrastructure to deploy it.
REST API design, auth systems, and database architecture built for production scale.
Multi-provider agent workflows, RAG pipelines, and self-hosted model deployment.
Containerized deployments and CI/CD pipelines for systems that need to stay up.
Live communication, queues, and connections to payments, travel, and messaging.
Click any card to flip it and see the challenge, my role, and the impact.
AI-Powered CRM Platform
Sales teams were losing deals to slow follow-ups — manually summarizing email threads and figuring out what to send next ate hours each week.
Sole backend architect — data models, API design, and the AI integration layer end-to-end.
Self-hosted Qwen 2.5 on RunPod cut per-request AI costs vs. pure API calls while Claude/Groq handled higher-stakes summarization.
Omnichannel AI Assistant Platform
Companies needed one AI assistant that stayed accurate to their own docs across three different channels, for both logged-in and anonymous users.
Designed the RAG pipeline — document chunking, S3-based knowledge storage, and the retrieval logic feeding the LLM.
One knowledge base, three channels, consistent organization-aware answers instead of three separate bots.
Payment Processing Platform
Reconciling transactions across two payment processors without manual spreadsheet work, while keeping financial records audit-ready.
Architected the backend end-to-end — gateway integration, reconciliation logic, and reporting.
Live in production serving 500–1,000 monthly users with automated transaction tracking.
Enterprise Travel Booking Platform
Six different supplier APIs (Sabre, Distribusion, Trenitalia, Italo, CDS Groupe, RateHawk), each with its own data format, needed to feel like one system.
Built the supplier sync layer, unified data model, and queue-based background processing.
Deployed live with SSL/DNS management — one consistent booking flow across all six suppliers.
Smart Android App
Farmers often lack quick access to plant pathology expertise, leading to delayed treatment and crop loss.
Trained the TensorFlow model, built OpenCV preprocessing, and integrated it with a Java Android frontend + Python backend.
90%+ model accuracy on disease identification, giving farmers actionable insights in the field.
From machine learning fundamentals to production backend systems.
Appnox Technologies Pvt. Ltd.
Internship
B.Tech, Computer Science Engineering (AI & ML) · CGPA: 8.086