Building agentic automation for D2C brands
Architecting end-to-end AI systems at Better Bright abandoned cart recovery, content pipelines, and multi-node agents connecting Shopify, messaging APIs, and analytics dashboards.
Currently at Better Bright (Wolf Pack Labs) — open to opportunities
Code. Build. Ship.
Building systems that work from distributed task queues and TCP socket programming to AI pipelines and agentic automation. I write code that runs, ship products that work, and think in systems.
0
Systems Built
0
Users Validated
0
Startups Shipped
01 / Now
Currently
A live snapshot of what I'm building right now: agentic systems, AI pipelines, and real products that serve real users.
Updated Jun 2026Distributed systems & LLM fine-tuning
Deep-diving into distributed task queues, TCP socket programming, concurrent processing patterns, and fine-tuning LLaMA for domain-specific AI applications.
Log > building agentic pipelines with n8n/
Log > fine-tuning LLaMA on Google Colab/
Log > shipping distributed task queues/
Log > designing AI content workflows/
Log > status: deep-work-mode/
Log > building agentic pipelines with n8n/
Log > fine-tuning LLaMA on Google Colab/
Log > shipping distributed task queues/
Log > designing AI content workflows/
Log > status: deep-work-mode/
02 / Work
Experience
Building real products at real companies. Three that say the most.
🤖
AI & Automation Engineer building agentic systems for D2C
End-to-end agentic automation using n8n and Google Opal abandoned cart recovery, AI content pipelines, Shopify webhook integrations.
🚀
Founded a tech-tourism platform from 0 to 1
Full product architecture, 200+ validated user sign-ups from a structured trial wave. Data-backed decisions to continue building.
🥥
Growth & operations for a D2C sparkling beverage brand
E-commerce ops, AI content pipelines, CRM segmentation, and creator management across platforms.
03 / Projects
Lab
Where I build things that don't have a deadline.
Near Deploy
Generative AI
SuitGen — AI Suit Pattern Generator
Fine-tuned Stable Diffusion model for Indian ethnic wear design. Culturally accurate, customisable pattern outputs.
→
Near Deploy
Generative AI
SuitGen — AI Suit Pattern Generator
Fine-tuned Stable Diffusion model for Indian ethnic wear design.
Deployed
AI / Education
Mentorix — AI Study Notes Generator
Government-recognised AI education platform. Presented to Director General of NIELIT, MeitY.
→
Deployed
AI / Education
Mentorix — AI Study Notes Generator
Government-recognised. Presented to DG NIELIT, MeitY.
Built
Systems
PyKV -> TCP Key-Value Store
Redis-inspired in-memory KV store in pure Python. Raw TCP sockets, LRU eviction, WAL crash recovery.
→
Built
Systems
PyKV -> TCP Key-Value Store
Redis-inspired KV store. TCP sockets, LRU eviction, WAL.
Some projects become products. Some stay as learning exercises. All of them teach me something useful about systems.
04 / Writing
From the archive
Why I Built a Redis Clone in Pure Python
TCP socket programming, LRU eviction, and Write-Ahead Logs — what building PyKV from scratch taught me about systems design.
Fine-Tuning Image Models for Niche Domains
How I fine-tuned Stable Diffusion for Indian ethnic wear design — a practical guide to domain-specific generative AI.
2 articles · Systems, AI, Building
All writing ↗
05 / Life
Outside the code
♟️
The best products and the best chess games share the same principle — think three moves ahead, but be ready to throw the plan away.
More about what I do outside code ↗