DEVELOPER’S MIND. CREATOR’S ENERGY.

Hey, I’m Dipu.Esmattoha Mondal

I build software.
I decode the AI era.

Software developerAI journalistYouTuber

From the code behind a product to the story behind a breakthrough. I build, experiment, and bring you along.

6+Years turning ideas
into working software
Open to collaborations
MEET THE BUILDER DIPU / BUILD. CREATE.
Professional studio portrait of Dipu, Esmattoha Mondal
HUMAN. STILL THE BEST FEATURE.
ESMATTOHA MONDALCall me Dipu.
BASED IN INDIA · BUILDING EVERYWHERE
CURRENT OPERATING SYSTEM
Build boldly.Question everything.Share the good stuff.

01 / ENGINEER

Software that works.

The APIs, systems, and products behind the experience.

02 / AI JOURNALIST

The story behind the tech.

AI breakthroughs, decoded with a developer’s perspective.

03 / YOUTUBER

Curiosity. On camera.

Tools, tutorials, and experiments on Play with AI.

02 MY YOUTUBE CHANNEL

YouTube / @playwithai

AI, explained.
On YouTube.

Play with AI with Dipu

New AI tools. Real experiments. The stories behind the tech. Your shortcut to understanding what’s next.

Watch on YouTube

03 THE AI PLAYGROUND

Less “what if.”
More “let’s build it.”

Small experiments. Big rabbit holes.
This is where curiosity becomes code.

BUILD MODE: ALWAYS ON
3 experiments

Agents

Give AI a job.

An agent that investigates slow database queries and proposes smarter indexes.

TypeScriptMongoDBTool calling
Inside the experiment

The loop reads structured query diagnostics, identifies missing index patterns, and proposes changes for review. The interesting part: giving the agent useful tools and clear boundaries.

AI knowledge

Ask your documents.

Turn a pile of PDFs into a knowledge system you can actually talk to.

RAGEmbeddingsVector search
Inside the experiment

An exploration of document chunking, semantic retrieval, and grounded answers. Preserving headings and code boundaries gives the model more useful context than splitting text by size alone.

Automation

Retire the busywork.

Messy supplier catalogs in. Structured product data out. Let the workflow handle it.

PythonPDF parsingAI workflows
Inside the experiment

The pipeline extracts supplier data and normalizes inconsistent product names. AI handles the ambiguous naming; deterministic code handles the calculations.

04 MORE THAN A JOB TITLE

Serious about the craft.
Curious about everything.

Dipu to friends. Esmattoha on paper. I’ve spent 6+ years turning complicated business problems into software people rely on.

My work spans Node.js and TypeScript backends, MongoDB database architecture, SaaS products, and AI knowledge systems. These days, that same curiosity takes me into AI—and in front of the camera. I like understanding how things work, making them better, and bringing people along.

The longer story on LinkedIn

Build with intention.
Share what you learn. Stay human.

THE JOURNEY SO FAR2019 — NOW
Lunix POSBackend Engineer & Database Architect2023 — NOW

Scaling multi-domain point-of-sale platforms and high-throughput MongoDB architectures.

Astria DigitalBackend Engineer2023 — 2024

Engineering high-reliability cloud services for web and mobile applications.

Netsqure TechnologiesFull-Stack Developer2023

Bridging backend architectures with frontend user interfaces and enterprise CRM systems.

Tradelia TechnologiesBackend Developer2020 — 2022

TypeScript rewrites, queue-backed notifications, and e-commerce data modeling.

Foxflue NetworkBackend Developer2019 — 2020

Foundations of server-side architecture, HTTP lifecycles, and micro-frameworks.

MY EVERYDAY TOOLKIT
TypeScriptNode.jsMongoDBVueAWSAI / RAG

OPEN TO THE RIGHT OPPORTUNITIES

Your next big thing.
Let’s build it together.

Build a product·Collaborate on content·Talk about AI
Say hello