Allweone
Flagship AI workspace
AI workspace with chat, agents, document tools, rich editors, realtime collaboration, and visual canvases.

I'm a Full Stack Developer focused on modern AI application engineering: Vercel AI SDK, React, Next.js, LangGraph, LangChain, browser agents, document pipelines, and voice-first tools.
Recent work includes Allweone, a large AI workspace; a Plasmo browser automation extension; a Gemini-powered computer assistant; Docling/FastAPI document extraction services; Raycast extensions; Plate.js editor systems; and local speech-to-text desktop workflows.
I still build the full product surface around the AI layer: authentication, databases, realtime state, polished interfaces, deployment, observability, and the small details that make tools feel reliable.
Stack
Plasmo
LangChain
LangGraph
OpenClawWork
Flagship AI workspace
AI workspace with chat, agents, document tools, rich editors, realtime collaboration, and visual canvases.

Browser automation extension
Plasmo extension that inspects pages and performs browser tasks through Chrome APIs and an agent UI.

Voice-controlled Windows agent
Wake word detection, browser listening, Gemini intelligence, tool calls, and spoken responses.

Raycast file conversion
Convert images, audio, video, and documents from a selected file or clipboard source inside Raycast.

AI video creation pipeline
AI stories, Gemini TTS voiceover, stock footage, captions, and final video rendering, fully automated.

Python SDK
Async Python SDK for UploadThing: uploads, signed URLs, list/delete/rename, ACL updates, FastAPI integration.

Real-time speech dictation
Python desktop transcriber that uses Faster Whisper to type spoken words wherever the cursor is focused.

AI communication app
AI-assisted articulation using Google GenAI, LangChain, Zustand, and streaming AI UI patterns.
AI slide deck builder
Open-source Gamma-style app that generates outlines, editable slides, themes, images, and presentation mode.

Full-stack social platform
Live feeds, stories, media upload, realtime chat, audio/video calls, auth, profiles, and notifications.

AI hotel search
Search a destination and get AI-generated hotel recommendations with preference handling and visual results.

Ask questions from PDFs
Extract information from PDFs through a ChatGPT-style document conversation interface.

DSA learning tool
Visualize sorting algorithms with speed controls, array sizing, and real-time state display.
Service marketplace platform
Book experts and local services with scheduling, payments, chat, and Agora audio/video sessions.

Quick-send browser extension
Select any text on a page, pick a contact, and open WhatsApp with the message prefilled.

Document intelligence service
FastAPI microservice converting documents into structured JSON and Markdown with image and table extraction.

Python desktop assistant
Local voice assistant with SpeechRecognition, PyAudio, pyttsx3, tray integration, and Windows APIs.
A snapshot of what clients and collaborators have said after working together.
I requested Deepak to tweak a project and he did a great job at it. Perfectly listened to the provided feedback, gave suggestions where to improve and fast delivery. Very pleasant to work with him. Eager to learn new things and a knowledgeable dev.
What a pleasure to work with Deepak. Young dev who's clearly talented with a strong passion for your complete satisfaction and solving any issues that should arise. I'm looking forward to working with Deepak further โ appreciate all your efforts!
Deepak built our company's product catalog website and absolutely nailed it. The design was clean, the layout made our range easy to browse, and he just got it โ barely needed any back-and-forth. Fast, professional, and the end result speaks for itself.
Building ALLWEONE alongside Deepak has been outstanding. He brings real ownership to the product โ not just writing code but thinking through the architecture, the user experience, and the business logic together. The kind of technical co-founder you want in your corner.