Very Good Ventures’ cover photo
Very Good Ventures

Very Good Ventures

Software Development

New Canaan, CT 18,397 followers

The Global Leaders in Flutter. We Build the Multi-platform Experience Layer for the Agentic Enterprise.

About us

Very Good Ventures (VGV) is the leading Flutter development agency for every device. We lead cross-functional teams to design, build, and grow multi-platform applications—mobile, web, desktop, and embedded—from a single codebase. Important: Very Good Ventures will never conduct interviews solely via chat or request payment or financial information from candidates. Official communication only comes from @verygood.ventures email addresses. Since Flutter's earliest days, we've been at the forefront of the framework, architecting enterprise-grade solutions for the world's most recognized brands. Today, we combine deep technical expertise with product strategy, world-class design, and AI innovation to help organizations move faster and build smarter. Our services span the full product lifecycle: Application Development, Product Design, Product Strategy, Agentic AI Solutions, AI & Strategy Workshops, and Marketing & Growth. We don't just write code — we engineer outcomes. We're pioneering the next wave of product development by integrating agentic AI and Generative UI into the apps our clients build: intelligent, adaptive experiences powered by the same cross-platform efficiency Flutter enables. Trusted by Google, Toyota, Universal Studios, American Airlines, JSX, Betterment, Hamilton, Dutch Bros, and Trackhouse Racing. 135+ engineers and creators delivering across the US, Latin America, and Europe. What makes us different: → AI-native product thinking built into how we work → One codebase, every platform—no duplicate development → Enterprise architecture that holds up at scale → End-to-end delivery from strategy through post-launch growth → Deep Flutter specialization backed by Google partnership Whether you're launching a new app, modernizing a legacy platform, or exploring what AI can do for your product experience, VGV brings the expertise, speed, and alignment to get there. Let's build what's next.

Website
https://verygood.ventures
Industry
Software Development
Company size
51-200 employees
Headquarters
New Canaan, CT
Type
Privately Held
Founded
2018
Specialties
mobile apps, design, technology consulting, product management, product development, digital technology, flutter, dart, web apps, desktop apps, and app development

Locations

Employees at Very Good Ventures

Updates

  • 5️⃣ Five apps. 4️⃣ Four operating systems. 🥇 One #Flutter monorepo. Marcus Twichel built #VeryYummyCoffee ☕️ — a complete coffee shop tech stack — to show what happens when you stop thinking of Flutter as a mobile framework and start treating every screen as a surface. The cashier toggles Pumpkin Spice Latte to unavailable on the POS. The kiosk updates. The mobile app updates. The menu board updates. No per-app integration work. One 🎯🐸 #DartFrog server, one multiplexed WebSocket, shared models on both sides of the wire. The open-source repo is live. The architecture patterns — shared Blocs, domain-agnostic UI components, a single ThemeExtension across every surface — are the same ones we use on production client work at Very Good Ventures. This is what Flutter's multi-device story looks like when you actually build it. https://lnkd.in/ggVSSquG #Flutter #Dart #DartFrog #OpenSource #MultiDevice #Monorepo

  • Building on a decentralized protocol sounds simple until you're managing 15 simultaneous WebSocket connections, reconciling conflicting data from independent relays, and designing an onboarding flow around raw cryptographic key pairs. Our engineers Dominik Šimoník and Hugo Walbecq just published their Part 2 deep dive into the real client-side challenges of building on Nostr, the protocol behind a growing wave of censorship-resistant apps. The problems are structural, not incidental: connection lifecycle management, relay discovery with missing user preferences, eventually consistent data without a server, key management that makes "Sign in with Google" look like a miracle of UX. None of it is unsolvable. But you need to know what you're walking into. Read Part 2 of our Nostr series 👇 https://lnkd.in/g6sGZ7cf #Flutter #Dart #Nostr #Decentralized #OpenSource #SoftwareEngineering

  • We rebuilt our website verygood.ventures from scratch. Not because our old site was broken, but because we needed a site that could move at the speed our team, and AI, now operates. It used to take us days to share our thinking to the world: SME interviews, writing, approvals, design - this whole process now takes minutes. Here's an overview of what we did: - We migrated our site from Webflow to Astro to unlock a faster, more flexible, and future-proof architecture—built for performance and scale. - The shift enabled an AI-native approach, giving us more control over content, workflows, and how we integrate emerging tooling across the site. - Moving off a monolithic platform reduced constraints and positioned us to iterate faster, experiment more, and better align marketing + engineering. If you're thinking about making a similar move, this might be helpful 👇 https://lnkd.in/eEpZQthQ #astro #claude #webflow

    • No alternative text description for this image
  • Very Good Ventures reposted this

    Just wrapped another talk on Generative UI in front of hospitality leaders at #DigitalTravelConnect in beautiful San Diego. Hard to beat the Rancho Bernardo Inn as a backdrop. What struck me most in the room: whether a brand has already shipped basic AI experiences or is still figuring out where to start, everyone is chasing the same thing, AI that actually touches the guest. Not another backend investment that never makes it to the screen. The appetite is real. The groundwork is laid. What's missing is the last mile. That's exactly what we're building at Very Good Ventures with GenUI, and we're already working with some of the leading brands in hospitality and travel to change how AI improves the customer experience. If you're thinking about what comes after the chatbot, let's talk.

    • No alternative text description for this image
    • No alternative text description for this image
  • The Very Good CLI MCP Server exposes real development workflows to AI agents: project creation from production-ready templates, test execution with coverage enforcement, dependency management across monorepos, and automated license compliance checks. Your AI assistant in Claude, Cursor, VS Code, or Windsurf can now scaffold a Flutter app, run your full test suite, and verify licenses. All without leaving the conversation. This isn't "MCP in theory." This is what MCP unlocks in a serious Flutter project: collapsed feedback loops, composable tool chains, and the ability to ask your AI assistant to do something and have it actually get done. We shipped it early because the value is already real. Read the full post to understand the architecture, configure your editor, and see exactly what these four tools can do. #MCP #Flutter #DeveloperTools #VeryGoodCLI #Dart #AI https://lnkd.in/gpSV7WMk

  • Banks and credit card companies are investing heavily in AI. Account opening abandonment just hit 67%.                                                                        Those two facts shouldn't coexist. But they do - and the reason isn't the AI. It's where the AI stops. We call it the Last Mile Gap: the distance between what your systems know about a customer and what actually shows up on their screen. Your data platform knows their cash-flow patterns, life events, and intent signals. Your mobile app shows them a static menu. The intelligence never reaches the interface. To understand exactly where in the account opening journey this hurts most, VGV ran quantitative Jobs to Be Done research. We surveyed customers who were actively considering opening an account - and stopped. What we found: they weren't abandoning because the form was too long. They were leaving before they ever started it - because nothing convinced them the account was right for them. That's the highest-impact moment in the journey. And it's almost completely unserved by existing digital experiences. Generative UI closes that gap. Check out the research, the opportunity map, and what it looks like in action - link in the comments. #GenUI  #DigitalBanking  #BankingInnovation  #GenerativeAI #CustomerExperience #Fintech

  • 🎮🔥🦄 Our colleague Taylor Hunt just competed in the #FlameGameJam2026 alongside his brother Morgan Hunt, and the lessons they brought back apply to way more than game dev. Their game, #SuppressedIntelligence, is a Plague Inc.-inspired strategy game where you play as an LLM spreading AI dependency across the globe. Windows 95 aesthetic and all. But the real story is what happened on day three: no playable build, an over-scoped design with four sector stats and three upgrade trees, and a deadline that wasn't moving. They made the hard call: cut two stat systems and two upgrade trees to protect the core gameplay loop. The result? A tighter, more intentional game. Shipped with 8 minutes to spare. Three takeaways that hit different whether you're building games or products: ✂️ Scope discipline is a skill, not a compromise. The best teams don't do less. They identify the core loop early and build toward it. 🤝 Pre-build alignment has a real cost when skipped. Two days of misaligned assumptions in a 10-day sprint is a significant tax. In a longer product cycle, it compounds. 🕹️ Even brief user testing changes everything. A quick playtest round revealed the tutorial wasn't optional — it was the minimum viable experience. Read Taylor's full breakdown on the VGV blog: https://lnkd.in/gs4ycDhH #Flutter #FlameEngine #GameJam #ProductDevelopment #ShipIt #OpenSource

  • A Python pipeline that took 24 hours to process NASCAR race data. A Rust rewrite — in a language our engineer had never touched — delivered in 3 weeks. Now it runs in real time. Our Director of Concept Technology, Kyle Crouse was at the Trackhouse Racing pit box at Daytona watching their telemetry pipeline fall further behind with every lap. The math was right. Python just couldn't keep up with 22,000+ messages per second across 36 cars at 50Hz. So he did something that would've sounded reckless a year ago: rewrote the entire pipeline in #Rust using #ClaudeCode as his implementation partner — despite never having written a line of Rust. The results: - Processing time per tick: 200-400ms down to 0.04-0.08ms - 227 million messages over a 3.5-hour race, zero restarts - Real-time data delivery to Flutter apps via LiveKit at 50Hz But here's the part that matters more than the speed numbers — the hardest problems weren't about code generation. They were about architecture. When the #Databricks consumer crashed from memory growth, the fix wasn't tuning a config. It was rethinking how data was structured. When outputs couldn't keep pace, the answer was tearing apart the monolith into 9 independent binaries with NATS JetStream fan-out. AI wrote the first draft. Kyle owned what shipped. That's what we mean at Very Good Ventures when we talk about taste and trust — AI speed grounded in the engineering judgment to know what's right and the discipline to prove it works under real conditions. The pipeline runs live at Darlington this weekend. Read Kyle's full breakdown: https://lnkd.in/gEQZT2-U #Rust #NASCAR #AIEngineering #Flutter #RealTimeSystems #ClaudeCode #VeryGoodVentures

  • We just shipped the #VeryGoodAIFlutterPlugin! 🦄💯💙 It changes how teams apply #Flutter and #Dart engineering best practices at scale when working with AI. Here's the problem we've been solving: best practices are only useful if they're accessible when you need them. We've spent years building consumer-grade applications used by millions, learning which patterns actually scale across distributed teams, large codebases, and complex architectural decisions. Those battle-tested principles—accessibility, performance, documentation, testing, design patterns—are now embedded directly into #ClaudeCode. This isn't just autocomplete. The plugin applies judgment. It reviews your widget structure for accessibility compliance, catches rendering bottlenecks, generates intention-driven documentation, strengthens your test coverage, and guides you toward architectural patterns that have proven themselves at scale. The plugin is open source and available now. For teams already using Claude Code, setup takes minutes. For teams new to AI-assisted development, we've included integration guides and examples that show how to layer best practices into your existing workflow. We're excited about the momentum in AI-assisted engineering and Flutter's readiness for it. This is the first of more tools we're building to help teams work smarter with AI assistance—without losing the judgment that makes code great. Explore the blog post here: https://lnkd.in/gnh7dPB3 #VeryGoodVentures #Flutter #EngineeringExcellence #BestPractices #AIAssisted

Similar pages

Browse jobs

Funding

Very Good Ventures 2 total rounds

Last Round

Series A

US$ 3.4M

See more info on crunchbase