Spawner AI

*Nothing herein is financial advice. Everything herein is purely informational. Nothing is a recommendation to buy, sell, trade, or own securities. Please consult a licensed professional before trading or investing.

Objectives

1. Create a portfolio-first investment product
2. Bridge the knowledge gap for traders
3. Abstract cumbersome financial data into actionable insights

The pitch

Many established financial applications claim they are democratizing finance for all. But I think they (we!) can do better. Just because users can access the markets and make trades or sift through rows of endless data, it doesn’t mean it is accessible or user-friendly. It’s accessible in the same way a casino or loan shark is accessible. You can walk in and find your fix with $1,000. But the odds are almost always in favor of your counter-party.

In the long run, the house (almost) always wins.

Spawner AI was built because we believe a decent amount of these core issues were addressable through better UX. Not quite solved, but addressable. There’s a lot of borderline predatory behavior that could be simplified for the end user with a badge here, a warning label there, or a guard rail over here. That’s why at Spawner the approach was to go after much of what we felt was low hanging fruit in wealth management.

Going portfolio-first

At the highest level of the Spawner product is a portfolio-first trading ideology. This is a fairly big pivot from how several traders evaluate and take action on their stocks of choice. Our goal was to curve short-term discretionary trading towards a more growth oriented, wealth creating approach. This is challenging for a handful of reasons. Here are two:

Reason 1: Data Overload

Users are very accustomed to loud interfaces that push the single-stock narrative. You’ve seen it. The rows and rows of metrics and numbers, graphs, green and red alarms, 10s of menus inside other menus, and on and on.

There’s nothing intrinsically wrong with projecting data onto a screen if that data is (somehow) actionable. But if it’s not actionable you’re creating a data overload scenario that does little more than confuse your customer.

That’s why after a lot of internal discussions and conversations with colleagues, we decided to narrow the focus and hide a lot of this data behind a simple menu system. Think of each UX decision as the furthest level of abstraction we’re comfortable to take a piece of data or an action to build or manage a portfolio.

Our overview page focuses on our ethos: portfolio-first. We want users to evaluate their basket of holdings at every turn, whether it’s one or one hundred stocks. If a user wants more information about a specific stock they hold, they can click on the ‘Stocks’ menu and get the same high-level insights.

Reason 2: Difficulty building portfolios

A clear outcome of the User Research we conducted before building Spawner, traders didn't seem to takeadvantage of portfolio-first thinking when deciding what stock (or basket of stocks) to purchase. In fact, many of the strategies shared werea little scary and haphazard. We even noted that we fall prey to these activities regularly because of scrolling news feeds, analyst ratings, and investment notes that mess with our psychology.

I want to clarify that none of these methods are “wrong”. People should be able to choose what they want to trade in their own way. But if a design system exists that can place you within tried and true guide-rails with powerful insights to help, wouldn’t you take it?

That’s a lot easier said than done, and few systems exist that provide this type of portfolio-building functionality. And ones that do seem to have potentially gone too far as to over-abstract information from the user. Balance is key. So we decided to take a stab at building it ourselves.

Here, users can build by choosing a basket of stocks, buy / sell individual stocks, rebalance existing portfolios, construct portfolios from ETFs, go with a pre-built portfolio (with historical performance embedded), or liquidate all of their holdings. By choosing a path, you’re immediately guided through a seamless workflow to get you a balanced portfolio in minutes.

Here's the modal you see after clicking the "Build Basket" radio button.

You’re then asked to make key decisions, including allocation amount and parameters that our bot (discussed later) can help you with.

Once you've made your selections, you're taken to the final stages of portfolio approval which are shown for any of the "portfolio building" options on the Spawner AI application. It consists of a chart highlighting the makeup of the portfolio, and a chart showing past performance.

Alternatively, what if you haven't done enough research on what makes a good portfolio? The customer might know they want to invest in a specific sector, and they're bought into the portfolio-first approach, but where do you start? That's where our handy pre-built portofolios come into play. Instead of choosing stocks independently, Spawner provides a list of pre-built portfolios indicating the type of holdings, approach, past return metrics, Sharpe ratio, and how risky historical performance was.

It should be noted during this process we've made sure to include a handful of reminders / warning systems to keep users on track and notified of any potential hazards.

With your portfolio system chosen, the powerful integrated ML builds you a balanced portfolio following the criteria you’ve provided. It’s that easy! One of my favorite analogies for what our build feature can do.

Imagine you receive a large box of Legos and someone asks that you recreate the USS Constitution from memory. Impossible, right? Now imagine having a guided blueprint describing the various sections that make up the ship, how many parts you have, and a general walk-through of how to build it. That’s Spawner.

‍

But what about all the other information traders want to know? Pricing, fundamental insights, news, sentiment, and everything else? We built something for that too.

Ask Wall Street (AskWST)

Our financial insights bot serves as your go-to for any questions or insights into the specifics of your portfolio. You can use it to: find sentiment, rebalance a portfolio, buy and sell the stocks of your choice, and so much more. We added a quick-action menu since memorizing commands can get tedious.

The best part about the bot is that it provides you around-the-clock alerts and insights about your portfolio. You can ask it questions like:

> Is my portfolio getting too volatile?
> What's Microsoft's current sentiment?
> Price of AAPL?
> 24h-change S&P
> Top earners today

Since the bot connects your portfolio to the external world, we had to consider the gravity of the UX behind this feature. That’s why we made the decision to keep it front and right-aligned at all times. The powerful NLP capabilities mean you can ask it most trading-related questions in plain English, which also means your research never stops. The power of having such an intuitive piece of technology at your fingertips while you build and test countless paper portfolios is what breaks the convention.

In fact, during development, we started to build a mobile version that would give a more seamless UX, enabling voice-to-text to ask financial questions on the go. Of course, ChatGPT has since taken stride but here's an initial mockup we were designing around.

Putting it all together

Spawner aims to prioritize the data that needs to be seen and demonstrate clear steps to retrieve additional information when needed. Different industries disrupt the UX conventions what seems like every year. Task management applications like Notion, Monday, and Asana are constantly pushing the needle by experimenting with new behaviors, or introducing existing features in different ways. Finance requires a certain set of expectations — no one wants to connect to a brokerage account and not know where to immediately find their portfolio’s total cash value. But we don’t think that’s doing justice to a lot of users who now have a wider breadth of digital experience and are ready to experience the benefits of an efficiently abstracted UX.