Meet Scout: Building a Trading Bot You Can Actually Follow

From Market Scanner to a Bot You Can Actually Follow

By Steven Zapf | Founder, Zapf Technologies LLC | Creator of Boom Stalker

For most of Boom Stalker's development, I was focused on finding opportunities.

Momentum. Volume. Relative strength. Trend. Risk. Market conditions.

The goal was simple: take a pile of market information and organize it into something a normal person with a job could actually use.

That worked.

But eventually I started asking a different question:

What happens after the signal?

Finding a stock is only one piece of the process. Someone still has to decide when to enter, how much capital to use, where risk should be managed, when profits should be taken, and when the trade is over.

And if Boom Stalker was ever going to become more than another stock scanner, those decisions needed to become part of the product too.

That is where Scout came from.

Meet Scout

Scout is becoming the face of the new Boom Stalker experience.

The original Boom Stalker workflow was heavily focused on scans and reports. The system would analyze market data, score opportunities, and surface stocks worth researching.

That was useful, but I realized something while building it: most people probably don't want another giant dashboard full of indicators.

They want to know what the system is doing.

Did it enter anything? What does it own? Why is it still holding? Did it take profit? Did it stop out? How much capital is deployed? How has it performed?

Those are much easier questions to understand.

So instead of making the user interpret every piece of the machinery underneath Boom Stalker, I started moving the product toward something much more intuitive.

Let the system do the work, and let the user follow along.

Scout became the face of that idea.

Scout at Work

The newest version of Boom Stalker has a simple concept at the center of the experience: Scout at Work.

When a user logs into a simulated portfolio, they can see what Scout is doing.

Open positions. Entry prices. Current prices. Position values. Realized and unrealized profit or loss. Trade history. Strategy activity. Portfolio performance.

Instead of receiving a scan and wondering what to do with it, the goal is to make the entire process visible.

You should be able to open your phone and immediately understand what happened.

That might sound obvious, but getting there required rebuilding a surprising amount of Boom Stalker.

The backend now has to understand the lifecycle of a trade. An entry becomes an open position. A target can trigger a partial exit. A stop can close what remains. Trades need timestamps, quantities, realized profit and loss, and a permanent history.

The product has to remember what happened even when nobody is watching it.

That is a much different system than simply saying, 'Here are five stocks that scored well today.'

And I think it is a much better direction.

Why Paper Trading Comes First

Right now Scout is operating in a simulated environment.

No real client money is being traded.

That is intentional.

Before anything involving real money should even be considered, I want to know exactly how this system behaves.

Not how it behaves in a spreadsheet.

Not how it behaves during a backtest.

How it behaves when the market is actually open and the entire chain is running.

Market data comes in. The strategy evaluates it. Scout makes a decision. The backend records it. The portfolio updates. The client sees it.

Then we do it again. And again. And again.

There will be winning days. There will absolutely be losing days.

That's the point of this stage.

I don't want to hide the ugly trades while showing the good ones. I want enough real testing data to figure out what actually works, what doesn't work, and what needs changed.

We just completed another live paper-testing session and got to see the new system finish green.

That's fun. But one green day doesn't prove anything.

The real work is building enough history to determine whether there is actually an edge.

Building the Product I Would Want to Use

One of the biggest lessons from building Boom Stalker has been realizing how different a developer's view of a product can be from a user's view.

I might care that a data pipeline completed correctly.

A user probably doesn't.

They care that they can open the app and see: Scout bought this. Scout sold half here. Scout stopped out here. Your simulated portfolio is currently worth this.

That's the experience I'm trying to build.

The complicated part should happen underneath. The simple part should happen on the screen.

And Scout gives that system a personality.

He isn't just a mascot sitting in the corner of the app.

The long-term idea is for Scout to become the guide through the entire Boom Stalker experience: account setup, risk questions, portfolio activity, trade explanations, alerts, and performance.

Eventually, I want someone who knows almost nothing about the system underneath Boom Stalker to still feel like they understand what their bot is doing.

Boom Stalker Is Changing

This also means parts of the original Boom Stalker are going away.

That is normal.

The product started with market scans because market scans were what I needed when I began building it.

But software gets better when you're willing to kill features that no longer fit the destination.

Boom Stalker isn't trying to become the world's most complicated stock scanner.

I'm much more interested in building a system that connects the entire process:

market data → strategy → decision → trade → portfolio → user

Scout sits right in the middle of that.

And for the first time, the frontend is starting to look like the backend I've been building for months.

That feels like a pretty major milestone.

Still Building

There is plenty left to do.

The mobile app is still being tested. The trading strategies still need far more data. The client experience will continue changing. Scout will get smarter. The portfolio system will get cleaner.

And I'm sure I'll break something along the way.

Probably several things.

But that's also why I started documenting this process publicly.

Boom Stalker didn't appear as a finished fintech product.

It started as an idea.

Then a spreadsheet.

Then some code.

Then an automated scanner.

Then a backend.

Then a client portal.

Now there's an app, a simulated portfolio, and a little green robot named Scout stalking the market.

Not bad for something that started with me wondering whether I could build a better way to organize stock data.

We're still early.

But Scout is officially at work.

Follow the Build

Boom Stalker is currently developing and testing its simulated trading platform. Follow the blog for product updates, trading education, testing results, and the continued development of Scout.

Boom Stalker is currently a development and educational platform. Simulated trading results do not represent actual client performance, and nothing on Boom Stalker should be considered individualized investment advice.