Inspiration
Inspired by my own struggles as an indie developer, I'm great at building apps but hate the marketing part. I think many solopreneurs face the same issue: they pour energy into creation but dread the grind that comes with telling people about it. A lot of these founders try to create TikTok videos but most fail to get significant reach even if they have a great product.
What it does
TokTraction allows content creators to analyze videos with a skill set they often do not possess so that they can create great content with a potential for going viral. In addition to basic analysis TokTraction simplifies content marketing for indie devs and solopreneurs when they create a campaign (e.g., views, followers, sales, signups) select a platform to market on, and describe their niche. The model generates tailored strategies with video concepts, scripts, shot lists, and storyboard prompts. Refine them via chat for dynamic updates, upload videos for analysis (alignment, predictions, suggestions), and iterate by inputting performance metrics to evolve the campaign. It's not fully autonomous yet but focuses on practical iteration to drive real results like growth and conversions.
How I built it
TokTraction is built with a full-stack Dart ecosystem using Flutter for cross-platform mobile and web apps, allowing me to ship the first version in just 9 days and iterate the current one in a couple of weeks. The AI is powered by mixed models from the Google Gemini series, and OpenAI gpt-4.1 series. The backend is deployed on Google Cloud Platform using their containerized Cloud Run product. The database is Firebase. Analytics via Amplitude. CI/CD via CodeMagic. Shorebird for shipping fast updates, keeps the web app and mobile apps more consistently aligned. Obviously RevenueCat for payments, was really great to have the recent updates for payments via Flutter web.
Challenges I ran into
The biggest hurdle will likely be marketing the marketing tool itself, ironic, right? At least it will create a good opportunity to validate the feature set. Development-wise, balancing AI autonomy without overpromising is tricky; it's not fully agentic yet, so I emphasized iteration over automation. Pricing is a little difficult in this space so I am experimenting with a variety of options. Moving forward, I'm surveying users for features like video and audio generation, ai-editing, etc. Ensuring I build what they need, not just what I think will be fun to build.
Accomplishments that I am proud of
Building a multi-platform app in under a month along with the backend all in one repo highlights the efficiency of Flutter/Dart. Empowering fellow indie makers with a that turns marketing dread into actionable wins feels like a win.
What I learned
So far I discovered the power of mixing models for balanced quality and cost. I was really happy with the current state of Flutter on the web, still found a couple pit falls but overall very pleased.
What's next for TokTraction
Next up, surveying users to prioritize features like auto-video editing, posting, and social platform integrations for true agentic autonomy. Figuring out a price point that resonates with my target audience. Long-term, building out a feature set that would make TokTraction an interesting product to startups that are a bit more mature.
Log in or sign up for Devpost to join the conversation.