Skip to content
View benbogart's full-sized avatar

Highlights

  • Pro

Block or report benbogart

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
benbogart/README.md

Ben Bogart Header Image

Ben Bogart

I build data and AI systems in hard domains — climate, carbon markets, agriculture at scale. Currently: building a personal multi-agent AI system for knowledge work, and developing a music-based typing trainer at the intersection of music and machine learning.

Looking for: founding engineer roles at AI startups, and director-level engineering or data leadership at climate-tech companies.


What I'm Building

AI agent system for knowledge work — A personal software factory: specialized agents with defined roles — product, research, implementation, review — that coordinate to help me build reliable software quickly. Backed by a structured knowledge repository, so agents draw on project context and accumulated decisions, not just the internet and training data.

keybr-music — A fork of keybr that I'm developing into a music-based typing trainer. My own concept, developed using my software factory — delegated automation with human-in-the-loop where it matters.


Featured Projects

Bird Vocalization Classification — TensorFlow classifier that identifies bird species from audio recordings. Built a live dashboard. The audio classification problem is more interesting than it looks: variable-length signals, overlapping calls, long-tail class distributions.

emissions-bulb — Real-time CO₂ emissions data from the Sense Energy Monitor API, visualized as a smart bulb. Climate data at the household level.

geospatial-python — Geospatial analysis and visualization in Python. Relevant to land-use, forestry, and carbon project work.

TickTickToOmniFocus4 — OmniFocus plugin that imports TickTick CSV exports. Small utility, but I needed it, so I built it.


Writing

I write about applied data science, ML engineering, and building diverse technical teams on Medium — published in Towards Data Science, Analytics Vidhya, and The Startup.

Worth reading:


Tech

Python · SQL · Databricks · AWS · Azure · Docker · Git · LLMs / AI agents · Geospatial (GDAL, GeoPandas)


Contact

linkedin.com/in/benbogart · ben@benbogart.com · benbogart.medium.com

Pinned Loading

  1. bird_vocalization_classification bird_vocalization_classification Public

    This project aims to build a bird species classifier that uses audio recordings of bird sounds (calls and songs) as its source material.

    Jupyter Notebook 5 1

  2. Pump-it-Up-Data-Mining-the-Water-Table Pump-it-Up-Data-Mining-the-Water-Table Public

    This project aims to use publicly available data to anticipate when a Tanzanian water well needs repair, ideally before it breaks so the pump can be repaired before it disrupts the local water supply.

    Jupyter Notebook

  3. waste-classification waste-classification Public

    The project is an exploration of using deep neural networks for classification of waste images using the Azure's python sdk and the resources of the Microsoft Azure Machine Learning Studio.

    Jupyter Notebook 2 1

  4. getting_started_with_azure_for_ml getting_started_with_azure_for_ml Public

    A notebook accompanying the medium article Getting Started With Azure for Machine Learning

    Jupyter Notebook 3 1