
COLE A HARTMAN
コール・ハートマン
Fixated on adapting software for humans and agents. I enjoy forward deploying to customers and solving their problems, turning technically ambiguous goals from idea to product with high agency and craft.
I spike in identifying key technical invariants, communicating cross functionally, and building novel solutions that will stand the test of time and scale.
You can find me on LinkedIn or reach me via email. I’m always open to chat.

WORK
Software Engineer Intern at Notion
Worked on various projects across AI Product and Desktop Infrastructure.
Summer 2026- Built a cloud-hosted AI agent harness to automate native tests, triage bugs, and validate releases on Desktop.
- Designed Electron APIs to write Notion pages directly to users’ filesystems, enabling local AI agent integrations.
- Secured the renderer-to-main process boundary with runtime validation, path containment, and atomic writes.
- Built native Markdown file handling for Notion Desktop, enabling 2M+ file opens since launch.
Open Markdown demo
Skill API demo
- Built a new Personal Access Token permission selector to make authentication more intuitive and secure. Try it now
- Implemented Deep Linking for fine-grained personal access tokens, allowing users to pre-fill token creation forms via URL. Try it now
Deep Linking demo
Updated permissions UI
Read the GitHub changelog ↗Software Engineer Intern at Converge
Maintained and improved the Electron desktop app, increasing engagement by 48% and reducing deployment time by 70%.
Fall 2024Software Engineer Intern at Happy Kiln
Led development of the modern e-commerce storefront using Next.js and Shopify’s Storefront API.
Summer 2024Software Engineer Intern at Pure Mentorship
Modernized company website and improved codebase structure within a 4-person engineering team.
Spring 2024PROJECTS
Engineered a low-latency brainwave pipeline to compute stress biomarkers and detect overstimulation with LLM-based signal analysis.
OSHA safety violation detection using Google Gemini vision + Supabase. Achieved 96% accuracy across 500+ evaluated site images.
Self hosted cloud service

BOOKMARKS
by Vicent Martí. published 2026.
by Jake Teton-Landis. published 2021.
by OpenAI. published 2025.