Projects I built

With code, design and fun

Here, I'd like to share some of my personal projects I've worked on, which included tasks such as creating vector graphics, analyzing data, doing web development, and building machine learning models, in different areas, like earth sciences, cryptocurrency markets, and marketing.

---

WhatLang

In brief, it's a machine learning model for programming language detection (80 languages supported). It is not based on a large language model or a BERT-like model. I preferred to follow a lean and optimized approach to have fast training and response times. The model is available as a Python library. Only 11 Mb! This could be useful for IDEs that want to identify and use the appropriate code syntax highlighter for the corresponding language. Feel free to test at whatlang.app.

---

Paperhand

Paperhand is a crypto dashboard with useful metrics that are not typically found on aggregators like Coingecko or Coinmarketcap. The platform gathers raw data from various sources to calculate indexes such as price similarity, volatility, consolidation, and more. These indexes are very helpful for crypto investors who need a tool to evaluate potential assets that may be undervalued or overvalued at some point. Explore the dashboard at paperhand.app. Built with Flask and Nim.

---

Pitchcast

I've always envisioned a better presentation format than video for pitches. Videos lack searchability, require retakes for mistakes, and don't support copy/paste or clickable content. That's why I created a proof-of-concept: synchronized slides with voice narration, clickable links, and a much lighter format than traditional videos. Another advantage is that if you need to update a text, logo, link, etc. You don't need to record again! Here is a demo: frankmoshe.com/pitch