Our Team
We built Agent GRaiL because we wanted it. Two founders who spent years grading cards by hand before building the AI-powered tools to do it better.
Why we built this
Every card collector has been there: you're looking at a raw card on eBay, trying to decide whether it's a PSA 10 candidate or a soft 8 that looks better in photos than it is. You're squinting at the corners, checking the centering ratio, wondering if that's a scratch or a print line on the surface.
We spent years doing that manually. Evan grades cards by hand. Donnie built a custom photography rig to capture the things standard photos hide. At some point we asked: why can't AI do this?
Agent GRaiL is the answer. A platform that brings professional-grade analysis to every eBay listing — not just the slabs you can afford to have graded.
— Donnie Lauring & Evan Barnard, Founders
Former professional athlete who spent years grading cards by hand before building the tools he always wished existed. Designed and built the lightbox-v4 photography rig for controlled card imaging, trained the CLIP-based grading models on real graded slabs, and architected the waterfall identity system that powers the catalog.
- →Designed lightbox-v4: angled raking-light rig that exposes surface defects invisible to standard phone photos
- →Built the CLIP ViT-B/32 + gradient-boosted grail head (BUY/PASS/REVIEW) classifier pipeline
- →Identified and documented the 18-sport canonical taxonomy used across the platform
- →PSA Data License holder (§8 attribution, §14 investment disclaimer wired into every catalog template)
Donnie Lauring
Founder & CEO
A lifelong sports card collector with deep hands-on grading experience. Evan brings the physical-grading perspective that keeps the AI honest — he's the first to spot when the model is overconfident on a surface defect the photo can't capture. His hardware engineering background shaped the lightbox work and the training data quality bar.
- →Hands-on PSA/BGS grader: personally submitted and tracked hundreds of raw cards through the grading process
- →Hardware engineering background: contributed to lightbox-v4 design and raking-light calibration
- →Product perspective: surfaces real collector pain points that drive the feature roadmap
- →Original beta tester: the model's early harsh-on-raw-cards bias was caught and corrected by Evan's feedback
Evan Barnard
Co-Founder
Our Methodology →
How we identify, price, and AI-grade cards — including known limitations.
Try AI Grading Free →
Search any card on eBay and get an instant BUY / PASS / REVIEW verdict.