A selected index of my experience and side projects.

Hi! I'm Dmitry Kuznetsov, a Product Manager working on AI integrations, platform upgrades, and UX improvements. Over the last decade, I've scaled a FinTech app to 1M+ MAU, picked up a Red Dot Award along the way, and helped upgrade platforms for groups like Orange and T-Mobile.

This page serves as a straightforward overview of my professional background, the AI-focused side projects I've been building recently, and some personal rabbit holes I explore.

I'm based in Vienna, Austria, bringing generative AI into rigorous, high-stakes workflows for legal professionals. Our new AI features are seeing 60%+ premium adoption and 12% month-over-month growth. We are building a true multimodal AI assistant.

Prior to this, I co-founded LEAV, earned my MBA in Entrepreneurship & Innovation from the WU Executive Academy, and completed the Stanford Product Innovation & Design program.

• Web AI
  Drafter A smart AI assistant built in a single day to simplify writing posts for a wellness brand. It analyzes the unique writing style, researches content ideas, and uses custom PDFs/URLs as a knowledge base. It generates new Instagram posts, carousels, and stories that are indistinguishable from the creator's own writing.
  
  Tech Stack: HTML5, CSS3, Vanilla JS, Vercel Serverless Functions, OpenAI API.
• Web Speech
  Lex — Legal Voice Platform A browser-based legal dictation and document drafting tool. Hands-free voice cursor control, custom voice commands, and a built-in library of standard legal clauses – allowing users to dictate, navigate, and edit documents entirely by voice. I built it as a personal sandbox to test and validate a few work-related ideas around legal workflow automation.
  
  Tech Stack: HTML5, CSS3, Vanilla JS, Web Speech API, LocalStorage.
• macOS
  DictationApp for Mac A native macOS menu bar application for AI-powered voice dictation (inspired by Wispr Flow), built in 2 days during an AI Hackathon. Triggered by a global hotkey, the app records user audio, transcribes it via Whisper, and refines the transcription using GPT-4o. The polished text is then automatically pasted directly into the active application's text field.
  
  Tech Stack: Swift, SwiftUI, OpenAI API, Swift Concurrency, AVFoundation, Carbon HotKey API.

  
• iOS
  AI Dictation Keyboard A smart iOS keyboard with AI-powered dictation, built in just 1 day during an AI Hackathon. Bypasses Apple's strict microphone limitations within keyboard extensions, executing real-time transcription and text refinement on the fly using OpenAI's Whisper API.
  
  Tech Stack: Swift, SwiftUI, UIKit, AVFoundation, App Groups (IPC), OpenAI Whisper API.

  
• Utility
  M4A → WAV Converter A local web utility for instant client-side conversion of iOS native M4A audio files to WAV format directly in the browser, without any server uploads. Developed as a personal tool for quickly preparing audio assets.
  
  Tech Stack: HTML5, CSS3, JavaScript (Web Audio API, File System Access API).
• Web
  This Portfolio Website A high-fidelity, premium single-page portfolio co-created in real-time with an AI coding agent. Built through structured agentic pair-programming, it implements a strict typographic scale, custom CSS variable tokens, responsive layouts, and a native HTML5 lightbox media gallery — demonstrating cutting-edge AI-assisted engineering workflows.
  
  Tech Stack: HTML5, Vanilla CSS3 (Design Tokens, Grid), Vanilla JS, Agentic AI (Antigravity).

• Endurance & Protocols
  Running & Biohacking I run half-marathons (with a full marathon on the horizon) and experiment with longevity protocols — keeping up with research. Enjoy cold and hot exposure. P.S. My father is bald, but when you see me, you'll see how much hair I have. If we get along, I'll share my secrets.
• Astrophysics
  Space & Galaxies Fascinated by the scale of the universe. I sometimes visit a local observatory outside Vienna and recently completed a course on galaxy evolution at Uni Wien just to understand how it all works.
• Technology
  AI Early Adoption Beyond work, I spend an unreasonable amount of time testing new AI tools, reading up on model architectures, and plugging experimental agents into my daily workflows the second they hit the timeline.