2026 — Now
Office Hours · Engagement Manager, Expert Data Lab
Advancing LLM capabilities in medicine and life sciences.
Alex Pryhoda
AI data & evaluationCurrently at Office Hours focused on expert data for training and evaluating LLMs in medicine and life sciences. Previously at Sepal AI, leading expert sourcing and human data delivery. Background in founding startups and machine learning research. Outside of AI, I roast specialty coffee. Based in New York City.
About
- Focused on building the best medicine and life sciences data to improve frontier models, and help AI accelerate scientific research.
- Experienced in building teams from 0 to 1, combining a technical background with hands-on management.
- Two-time co-founder: Insecterra, an AI-powered smart farming company ($1M+ raised), and ddd, a creative agency (20+ projects).
- BS in Computer Science, Ukrainian Catholic University. MS in Management, UCU Business School (LvBS).
Experience
AI Data & Evaluation
2025 — 2026
Sepal AI (acquired by Mercor) · Strategic Projects Lead
Human data, SFT, RLHF, evals and expert sourcing. Employee #8.
Founder
2019 — 2024
Insecterra · Co-founder & CEO
AI-powered smart farming. Took it from 0 to 1 and raised $1M+.
2020 — 2024
ddd · Co-founder & CTO
Full-cycle creative agency. Web and product development, 20+ projects. Forbes Ukraine Next 250.
Machine Learning
2018 — 2019
Biomedical Computer Vision Lab · Machine Learning Researcher
Institute of Computer Science, University of Tartu. Deep learning for biomedical imaging; PerkinElmer-funded research.
2017 — 2018
Ralabs · Machine Learning Engineer
Computer vision: object detection and segmentation for industrial waste sorting.
Projects
2025 — Now Synesthesia Small-batch specialty coffee roaster in New York.Publications
2019 Tissue Segmentation in Histopathological Whole-Slide Images with Deep Learning Ukrainian Catholic University. A novel deep learning pipeline for tissue segmentation in histopathological whole-slide images. 2018 Astrocytes Segmentation in Brain Microscopy Images University of Tartu. A deep learning approach for astrocyte segmentation in brain microscopy images.Contact
Always happy to talk about AI data, evals, startups, or coffee.