Hey! I’m Perry, a data scientist and machine learning engineer with experience leading data and machine learning initiatives in early-stage companies. I helped start Fishtail, a trade finance automation platform where we use a lot of alternative supply chain data and machine learning to automate back-office processes and de-risk global trade. I previously led data science at Farm Dog (acquired by Deveron), and have worked on a variety of projects ranging from medical imagery segmentation to algorithmic trading at a hedge fund.

Now

I currently live in Missoula, Montana. In my free time I love exploring remote wilderness areas by climbing, hiking, or snowboarding. Right now I’m interested in, researching and talking with folks in the following domains:

• Regenerative agriculture
• Wildfire mitigation
• Supply chain efficiency
• Human health optimization

Experience

• ML engineer and Co-founder at Fishtail
  • Built a document understanding engine to extract structured information from unstructured files of any format
  • Developed a reconciliation engine to automate trade finance repayments
  • Created a daily global maritime port health index to monitor maritime supply chain risk
• Lead Data Scientist at Farm Dog where I developed data pipelines and in-field tools to support agronomists with pest and disease data.
• Project Work
  • Built a highly configurable Python library to stress test in-hospital care software systems, simulating various hospital layouts and complex patient scenarios.
  • Conducted research in collaboration with a leading radiologist to develop an image segmentation model for identifying colorectal patients and detecting lymphoma tumors using medical imagery.
  • Optimized slow-performing database queries for a platform connecting EV battery metrics with home electricity rates, reducing query execution time significantly and enabling seamless calculation of charging costs and savings.
• Quant Analyst @ Strix Leviathan, a crypto hedge fund, where I researched and developed algorithmic trading strategies against a wide variety of cryptocurrencies.

Personal Projects

• Bird Scooter Nest Generator: Built a web-app that uses machine learning to optimize charging and redistribution locations for Bird scooters.
• Reverse Engineering the Walk Score Algorithm: Identified the key features used in the Walk Score algorithm to predict walkability for Seattle neighborhoods.
• Hiking Recommendation System: Developed a personalized recommendation engine for hiking trails. A sort of Netflix for hiking trails.

Tech Stack

Python, ML Libraries (NumPy, Scikit-learn, Pandas, PyTorch, fastai, TensorFlow, Keras, Turi Create), Data Viz Libraries (Matplotlib, Seaborn), Web App Frameworks (Django, Flask, FastAPI), LLMs (OpenAI, Anthropic, Llama, Groq), APIs, Azure, AWS (Batch, EC2, RDS, S3), GCP (BigQuery, Vertex AI, Cloud Run), Docker, Heroku, Git, Bash, JavaScript, MongoDB, PostgreSQL, Geographic Information Systems, and Go.

Contact

If you’re working on something cool or want to chat, feel free to reach out at perryrjohnson7[at]gmail.com.