Rune

The market doesn't beat traders—emotions do. Fear, greed, panic. You've felt it—that moment when logic disappears. Our AI catches those split seconds before they become expensive mistakes.

Comment

Inspiration

I identified a critical gap in the trading technology landscape:

While platforms like Tradezella dominate journaling,

They've been slow to integrate meaningful AI capabilities.

More importantly,

I noticed that trading psychology—

Despite being constantly discussed by educators—

Remains frustratingly vague and unactionable.

Here's what really got me fired up:

MOST traders already have winning strategies!

They're not failing because they don't know what to do—

They're failing because emotions hijack their decision-making under pressure.

That "aha moment" sparked the creation of Rune:

An AI coach that bridges the gap between knowing what to do and actually doing it when your heart's racing and money's on the line.

What it does

Rune transforms traditional trading journals into intelligent coaching systems that actually understand you.

Users upload their trading manual(playbook), and rune doesn't just read it—it deeply analyzes and internalizes every rule, preference, and strategy,

When traders submit chart screenshots and trade notes,

Rune becomes that brutally honest friend who cross-references their actions against their own rules, providing real-time validation and psychological insights that sting (in a good way).

To make it even more human to bring out the honest friend feeling;

We implemented Tavus

And it generates a video summary of your trades;

It tells you where you went wrong and where you went right and ways to improve based on YOUR data.

The platform identifies emotional patterns you didn't even know you had, validates setups with surgical precision, and delivers personalized performance reports that connect psychology directly to P&L outcomes.

It's like having a trading therapist who actually knows what they're talking about.

Wait theres more

And here is where it gets really powerful:

The journal doesn't just record your trades—

It learns from them.

Based on your trading history, win/loss patterns, and emotional states;

Rune generates actionable insights that help you make better decisions going forward.

It spots trends like "you're 73% more profitable when you trade during London session" or "your win rate drops to 23% when you're feeling 'anxious'"

And uses these insights to guide future trading decisions.

It's like having a trading therapist who actually knows what they're talking about and remembers every mistake you've ever made—

And uses that knowledge to make you better.(and maybe make you more money).

How We Built It

The magic happened on Bolt.new—and I mean magic.

This platform enabled rapid development of a beautiful, production-ready full-stack application that honestly blew my mind.

Bolt's intuitive interface and powerful backend capabilities allowed me to create a visually stunning platform with complex AI integrations—

Something that would have taken months in traditional development environments and probably driven me to question my life choices.(or loose a few hairs)

I developed 90% of the functionality directly on Bolt, creating a polished user experience with seamless TradingView integration and sophisticated chart analysis.

The development flow was so smooth it almost felt like cheating.

Only when our codebase expanded beyond 14,000 lines did we migrate final optimizations to Claude Code for deployment scaling.

Which to be honest is extreme just even mentioning it. But the fact that the platform could generate roughly 13,000 lines of code still fascinates me

Challenges we ran into

The primary technical challenge emerged when our codebase exceeded Bolt's compression capabilities at ~14,000 lines,

Forcing a platform migration mid-development.

Honestly, this was both frustrating and impressive—it actually speaks to Bolt's incredible power that we were able to build such a feature-rich application entirely within the platform before hitting any limits.

However claude code wasn't as easy to use; I had to set up supabase; sometimes I would have to manualy deploy the edge functions; other times I had to manually debug the supabase edge functions

And P.S; I know I'm sounding spoilt but before in bolt I could just press attempt fix and 9 times out of 10 the error would go soon afterwards.

And the migration to Claude Code brought about a similar hurdle: rate limits even on the basic pro plan;

Which created some hair-pulling moments during intensive development phases, Requiring careful resource management and the kind of optimization strategies that make you appreciate every SINGLE API call.

Accomplishments that we're proud of

We successfully built a beautiful, fully-functional AI trading coach that rivals commercial platforms in both aesthetics and functionality—all developed primarily on Bolt.new.

I'm genuinely proud of what we achieved here.

The platform performs multi-timeframe chart analysis directly from TradingView data,

Features an intuitive user interface that feels professional and polished,

And most importantly delivers personalized video summaries through Tavus integration that are actually engaging to watch.

Most importantly,

We created a complete full-stack application that actually works in production—no smoke and mirrors here.

With Bolt handling everything from frontend beauty to backend complexity seamlessly,

The result is a platform that looks like it was built by a seasoned development team,

Not a solo developer caffeinated beyond reason during a hackathon timeframe.🤣🤣

What we learned

Building a comprehensive full-stack application with AI integration demands unwavering commitment and precision,

But Bolt.new made the impossible feel achievable.

The platform's ability to handle complex frontend/backend integration while maintaining visual excellence accelerated development beyond what I thought was possible—seriously, it was a game-changer.

I also learned that AI assistance accelerates development but requires extremely specific instructions (it definitely cannot read your mind, despite what we hope).

The iterative process of translating complex trading psychology into actionable algorithms proved both challenging and deeply rewarding, made manageable by Bolt's seamless development environment that kept me sane throughout the process.

What's next for Rune

Our immediate focus involves comprehensive testing across different AI models and APIs to optimize chart analysis accuracy—because good enough isn't good enough when people's money is involved.

We're exploring the possibility of fine-tuning our own specialized model for trading psychology and chart pattern recognition, which gets me genuinely excited about the potential.

But we're not stopping there. T

he next major leap involves integrating advanced TradingView chart capabilities and enabling direct trading from within the platform itself.

Imagine this: you analyze a setup, get the AI's approval, execute the trade, and have everything automatically journaled without lifting a finger.

That's the kind of seamless experience that makes journaling effortless and actually sustainable.

Even more exciting is our plan to automatically import TradingView charts that users have already analyzed—complete with their drawings, indicators, and annotations.

Right now, having to manually upload screenshots feels clunky, and frankly,

It's a barrier we need to eliminate.

We want the platform to intelligently pull in your existing chart work and understand your analysis patterns without any manual intervention.

The roadmap also includes expanding beyond individual coaching to institutional applications—think hedge funds and prop trading firms that want to scale psychological discipline across their entire team.

The timeline and scope will ultimately depend on funding and market validation, but the solid foundation built on Bolt.new gives us confidence that scaling and adding these ambitious features will be as smooth as the initial development process—and that's saying something.

The goal is simple: make Rune so integrated into your trading workflow that it feels like a natural extension of your brain, not another tool you have to remember to use.

Built With

Share this project:

Updates