Sasha Denisov · Berlin
Twenty years of software.
Engineer, architect and builder across AI, Flutter and cloud. CTO @ Brainform.ai and Google Developer Expert. I have taken plenty of large products from an idea to real scale, and I am always happy to talk shop.
About
I have spent more than two decades architecting scalable systems and taking products from an idea to real scale — across fintech, digital media and entertainment.
I am a Google Developer Expert in AI, Cloud, Firebase, Flutter and Dart, and co-organize the Flutter Berlin community. My work spans the full stack of a modern product: backend and cloud architecture, cross-platform apps with Flutter (including Android TV and Apple TV), and AI — from cloud-scale generative systems to agents and fine-tuned models running on-device.
Whatever the stack, the through-line is the same: helping teams make the right architectural calls and ship products that last.
Products
Things I have built
These are the projects I can show publicly. Much of my work — building products inside big tech — stays under NDA.
CTO
AI shopping for e-commerce. An on-site agent that drives ~2× higher conversion, and an agentic layer that makes catalogs visible to personal AI agents like ChatGPT and Gemini. Works with Shopify, BigCommerce, Magento and WooCommerce.
Visit →
Creator & maintainer
An open-source Flutter plugin for on-device AI. Runs Gemma 4, Gemma3n (multimodal vision + audio), Qwen3, Phi-4, DeepSeek R1 and more — across Android, iOS, Web, macOS, Windows and Linux, with LoRA fine-tuning and function calling.
Visit →
Creator & maintainer
Flutter, now on Apple TV. An open-source, production-ready custom Flutter engine for tvOS — build TV apps with the framework you already know, native on the largest screen in the room.
Visit →What I help with
Consulting & advisory
AI / GenAI products
Strategy and launch of AI products — from a first prototype to production, with the right model and architecture choices.
AI agent orchestration
Designing agentic systems: multi-agent flows, MCP, smart routing — and keeping the path to production short.
On-device & hybrid AI
Running models on the device and routing intelligently between local and cloud: on-device inference, fine-tuning, multimodal, and hybrid pipelines that cut cost and protect privacy.
Flutter products
Architecture and delivery of cross-platform products on Flutter, built to scale.
Flutter for TV
Android TV & Apple TV with Flutter — a rare specialty, from custom engine builds to shipping TV apps.
Cloud & backend
Firebase, Genkit, serverless and scalable backend architecture for AI-powered apps.
Product launch 0→1
20+ years taking large products from an idea to scale — across fintech, digital media and entertainment.
Mobile & team strategy
Optimizing mobile strategy: tech choices, team structure and process for faster, measurable delivery.
Public speaking & community
Conference talks, workshops and community building — 30+ stages worldwide and counting.
Expertise
The toolkit
Open source
Selected projects
Run Gemma AI models locally, on-device, from a Flutter app.
Flutter on the big screen — Android TV and Apple TV.
A proof of concept for handling deep links properly in Flutter.
A speaker-rating app showcasing varied architectural patterns.
An AI conversation demo app.
Speaking
On stage
Building Offline AI Agent in Flutter
Shipping autonomous AI agents that run fully on-device, no backend required.
Hybrid AI in Flutter with Genkit — On-Device and Cloud Models
One pipeline that routes between local and cloud models with Firebase Genkit.
Evolution of flutter_gemma
From a simple wrapper to a cross-platform on-device AI engine.
Flutter and Apple TV. Story of one challenge
Building a custom Flutter engine to bring Flutter to Apple TV.
Future of Mobile AI
What on-device intelligence really means for app developers.
AI Agent Specialization: RAG vs Fine-tuning
Choosing the right approach to specialize an AI agent for your domain.
On stage · video
Talks on video
▶
Evolution of flutter_gemma
FlutterConf 2026
▶
Building Voice AI Agents That Listen, Understand, and Sell
GenAI Summit EU 2026 · Valencia
▶
Building Offline AI Agent in your Flutter app
Fluttercon EU 2025
▶
Optimization of mobile development strategy for business impact
LDX3 2025
▶
Building Next-Level Flutter Apps with AI using Firebase Genkit
Flutter Heroes 2025
▶
Flutter + Firebase: an explosive combination
CPH DevFest 2024
▶
Flutter on Apple TV
Full Stack Flutter 2024
▶
Shake the Flutter tree
Flutter Heroes 2023Writing
Articles
Hybrid AI in Flutter: Routing Between On-Device and Cloud Models
A practical pattern for deciding when to run locally and when to call the cloud.
Read → 2026On-Device RAG for App Developers: Embeddings, Vector Search, and Beyond
Building retrieval-augmented generation that runs entirely on the device.
Read → 2025On-Device Function Calling with FunctionGemma
Giving small on-device models reliable tool-use and function calling.
Read → 2025Fine-Tuning Gemma with LoRA for On-Device Inference
Separate LoRA weights for Android, iOS and Web on-device inference.
Read → 2025The Dawn of Offline AI Agents in Your Pocket
Why offline-first AI agents are the next shift for mobile.
Read → 2022Flutter for Apple TV
Bringing Flutter to the living room — building for Apple TV.
Read →Work with me
Book a consultation
Not sure yet? Start with a free 15-minute intro — tell me who you are and what you want to cover, and I will confirm a slot. Ready for a deep dive? Book the paid one-hour consultation.