cd /news/ai-products/nitai-dean-idp-software-author · home › topics › ai-products › article
[ARTICLE · art-146111] src=idp-software.com ↗ pub= topic=ai-products verified=true sentiment=· neutral

Nitai Dean: IDP Software Author

Nitai Dean is cofounder and chief technology officer of DocuPipe, a New York document extraction platform operated by Hoss Inc. that converts PDFs, scans, photos and spreadsheets into JSON matching a customer-defined schema, running as a hosted API or inside a customer's own cloud account or Kubernetes cluster. Dean leads the machine learning and engineering side alongside cofounder and CEO Uri Merhav, and previously ran the machine learning consultancy Hoss Technologies with Merhav from 2017, building models for startups in Silicon Valley, New York and Tel Aviv before the team turned to its own products, including the document parsing tool DocuPanda that became DocuPipe. Dean authored DocuPipe's comparison with Extend, benchmarking both products on DocuBench, a public set of 72 hard documents in 12 languages, and on Extend's RealDoc-Bench, publishing the labels, scorer and raw outputs on GitHub.

read1 min views3 publishedOct 6, 2026
Nitai Dean: IDP Software Author
Image: Idp-Software (auto-discovered)

Cofounder and CTO, DocuPipe · LinkedIn · docupipe.ai Nitai Dean is cofounder and chief technology officer of DocuPipe, a New York document extraction platform operated by Hoss Inc. DocuPipe turns PDFs, scans, photos and spreadsheets into JSON that follows a customer-defined schema, and runs either as a hosted API or inside a customer's own cloud account or Kubernetes cluster. Nitai leads the machine learning and engineering side of the product, alongside cofounder and CEO Uri Merhav.

Before DocuPipe, Nitai and Uri ran Hoss Technologies, a machine learning consultancy that from 2017 built models for startups in Silicon Valley, New York and Tel Aviv. In a 2024 essay, Nitai describes how the consultancy turned to its own products, first the image tools PhotoBear and the storytelling app MyTales, and then DocuPanda, the document parsing product now called DocuPipe, built because the parsing tools the team had used on client projects fell short. Nitai reached Master rank in Kaggle machine learning competitions and studied at Tel Aviv University.

Nitai wrote DocuPipe's comparison with Extend, which runs both products on DocuBench, a public set of 72 hard documents in 12 languages, and on Extend's own RealDoc-Bench, and publishes the labels, scorer and raw outputs on GitHub so others can rerun it.

Contributions #

  • DocuPipe Vendor Profile : Profile of DocuPipe's schema-based extraction platform, covering the parse, standardize, split, classify and review steps, credit pricing, the DocuBench results, hosting regions and certifications, the on-premises installation and vendor-reported customer outcomes.

Nitai joins other vendor-side IDP Software contributors who document platforms they build, among them Gael Anaya of Claix and Rayan Azmatally of ORKOM.

── more in #ai-products 4 stories · sorted by recency
── more on @nitai dean 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/nitai-dean-idp-softw…] indexed:0 read:1min 2026-10-06 · —