Weekly Roundup
Welcome to Weekly Roundup #2 at ForecastOS! We're glad to have you here, following our journey :)
This week I split the newsletter into two sections. The first section contains thoughts / musings relevant to ForecastOS. The second section contains what we worked on: product updates, UX improvements, partnerships, etc.
I hope you enjoy the new (experimental) format!
1. Musings: Growth Hacking ForecastOS
TL;DR: We are going to create a click through ForecastOS demo in a restricted environment from winning Kaggle solutions, link to it at the top of our landing page, share it on platforms like Hacker News, and capture the emails of users who want early access.
When you show the world something new, people take notice. I believe:
- A great demo for an original idea, like the v1 Rails demo from 2005, creates viral growth...
- ...if shared on a platform with sufficient interest and scale
Bonus points if the material in question creates healthy debate.
I witnessed how (mostly healthy) debate can create engagement first-hand after writing the article "YC Software Startups: Value and Initial Programming Language Used." The article created spirited debate over causality vs correlation, among other things, and has been read by tens of thousands of people.
It spent the better part of a day near the top of Hacker News:
It was retweeted by Matz, the creator of the Ruby programming language:
And it made the rounds on Reddit, Slack, and numerous other platforms. Pretty cool for an article about code! Even after the viral traffic slowed, a lasting effect remained; despite not updating that article for several years, it still receives a few hundred organic visitors per month!
How can similar viral growth be created for ForecastOS? I think it can be done by:
- sharing a public ForecastOS demo that anyone can access and click through without creating an account (in a restricted environment)...
- ...that visualizes and organizes winning, public Kaggle solutions from financial forecasting competitions in a format that can be understood by anyone non-technical with a passing interest in financial AI...
- ...and capturing the emails of those interested in obtaining early access!
My goals are to:
- Get +10k quality views by sharing the demo on relevant platforms (e.g. Hacker News). Spark healthy debate over the future of AI in the workplace and the (growing) involvement of less technical contributors / managers
- Demonstrate that anyone, including non-technical executives, can effortlessly understand, analyze, and manage AI-powered financial forecasts through an intuitive UI. Demonstrate that ForecastOS integrates seamlessly with existing machine learning pipelines. Demonstrate how forecasts would translate into real-world performance via backtests
- Showcase different forecasting use cases, like fraud, credit events, and (future) prices. Expand ForecastOS functionality
- Capture emails of +100 users interested in early access
- Reduce the friction to try ForecastOS to 0
Please stay tuned for the first (frictionless) ForecastOS demo in the coming weeks!
2. Work: What Got Done
ForecastOS is a new startup. As such, we try to spend 100% of our time either building or selling. We log that effort here.
This past week, as it pertains to selling, we:
- Continued the outreach process for our financial AI management offering; if you know anyone we should talk to, please let us know! We have capacity to onboard 1 additional POC (proof of concept) client over the next 3 months
This past week, as it pertains to building, we:
- Added InvestOS support for Python 3.8 (released ~4 years ago). Previously InvestOS only supported Python 3.9 and up
- Iterated on desired backtesting functionality with our first client
- Prepared to onboard our first full-time employee in several weeks!
A final, fun update: my sister got married last week! For those of you I saw in Toronto over the last few days - it was a pleasure!
Until next time!