MarkTechPost · Aug 2024
Third-party coverage of IRIS describing prompt-driven labeling, zero-shot model selection, and faster CV data annotation.
Company research
Public company, workplace, funding, and market signals
Updated Jul 29, 2026
Overview
Overeasy is a San Francisco-based AI/computer-vision startup in Y Combinator’s Summer 2024 batch building IRIS, an AI agent and open-source framework for auto-labeling visual data and orchestrating zero-shot vision workflows. Public materials identify Kelly Lu and Anirudh Rahul as the founders; I did not find a separate public CEO disclosure.
Primary product
IRIS, an AI agent/framework for auto-labeling visual data and orchestrating zero-shot computer vision models
Founded
2024
Headquarters
San Francisco, California, United States
Team size
1-10 employees
Industry
Technology, Information and Internet
Sub-industry
Business/Productivity Software
Offices
San Francisco, California, United States
Business model
open-source software
B2B software
developer tools
Funding
Stage
Early Stage VC
Total raised
$500K
Latest round
Early Stage VC · Feb 2026
Feb 2026
Jan 2024 · Y Combinator
People & investors
Investors
Work & culture
Research-heavy, builder-centric, and open-source oriented; public materials emphasize fast computer-vision workflows, annotation quality, and developer ease, but I found no formal public benefits or culture page.
Product & market
Differentiators
- —IRIS automates visual-data labeling from prompts and visual hints
- —Selects the best zero-shot models for a given use case
- —Supports iterative correction of annotations
- —One-click training and deployment of CV models
- —Open-source framework with an MIT license
Technology
HTML
Python
Gradio
CLIP
OwlV2
GroundingDINO
YOLOWorld
Playwright
Selenium
Competitors
Labelbox
Roboflow
CVAT
V7
Scale AI
Scale & contact
Estimated revenue
$100K-$250K (estimate)
Latest news
Y Combinator on LinkedIn · Aug 2024
YC’s launch post says Overeasy (YC S24) is building IRIS to distill large foundation models into faster, task-specific computer-vision models and accelerate labeling 100x.
Y Combinator · Aug 2024
Official launch write-up with product details, team background, and the claim that IRIS can auto-annotate massive datasets and train/deploy models with one click.
Hiretop · Aug 2024
Profile-style article on Overeasy, its founders Kelly Lu and Anirudh Rahul, and IRIS’s role in speeding up computer-vision model development.
Jobs and careers at Overeasy
0 jobs at Overeasy
No open positions
Check back later for new openings