The problem is a universal one in enterprise back offices: a stack of invoices, tax forms, or handwritten voting records arrives, and a team begins the slow, expensive process of manual digitization. The solution, for a growing number of developers in Latin America, is a single API call to a Chilean startup. DocuTray, founded in 2025, is building a developer-first platform to identify, extract, validate, and export structured data from PDFs, scans, and even manuscripts. Its early traction, anchored by processing hundreds of thousands of election pages, is a test of whether a region-specific approach to document automation can carve out a sustainable wedge.
The Wedge: One API Call, Five Steps
DocuTray’s pitch is technical simplicity for a complex workflow. Instead of stitching together separate services for optical character recognition (OCR), data validation, and systems integration, the company offers a unified API that promises to handle the entire pipeline. A developer sends a document; the system identifies its type, extracts the relevant data into structured JSON, validates it against configurable logical and numerical rules, enriches it by cross-referencing customer databases, and finally exports the result to a business system like an ERP [DocuTray]. The company claims this process takes less than 20 seconds per document, representing a time saving of over 90% compared to manual work [El Mercurio, Aug 2026]. For developers, the value proposition is a reduction in integration complexity and maintenance overhead.
Traction Through a Stress Test
The most compelling evidence for the platform’s scalability came not from a traditional enterprise contract, but from a public-sector stress test. In late 2025, DocuTray processed almost half a million pages associated with presidential election voting records in Chile and Honduras [El Mercurio, Aug 2026]. This involved handling handwritten tally sheets, a format that challenges many OCR systems. The project served as a large-scale validation of the infrastructure. In the twelve months leading up to August 2026, the company processed a total of 700,000 pages [El Mercurio, Aug 2026]. By July 2026, through early enterprise integrations, it had processed more than 350,000 documents comprising over 520,000 pages [Carabela Ventures, July 2026].
The Team Behind the Build
The founders are not first-time operators. CEO Matías Sánchez Cabrera and CTO Roberto Arce previously co-founded and scaled Datalized, a data consultancy they grew to a 25-person team, serving over 80 enterprise clients and generating $3.5 million in cumulative revenue [Carabela Ventures, July 2026]. Both have backgrounds as data scientists at regional tech companies like Cornershop and ComparaOnline, where Arce spent seven years leading predictive modeling systems [Carabela Ventures, July 2026]. This shared history building and servicing enterprise data pipelines in Latin America informs DocuTray’s product decisions, particularly its focus on validation and enrichment steps that address local business rules and data formats.
| Founder | Role | Key Prior Experience |
|---|---|---|
| Matías Sánchez Cabrera | Co-founder & CEO | Co-founded Datalized; Data Scientist at Cornershop & ComparaOnline [Carabela Ventures, July 2026][RocketReach] |
| Roberto Arce | Co-founder & CTO | Co-founded Datalized; led predictive modeling at ComparaOnline; roles at Cornershop & Datalized [Carabela Ventures, July 2026][DocuTray] |
Funding and the Path to Mexico
In August 2026, DocuTray closed a $300,000 pre-seed round led by Mexican venture firm Carabela VC [El Mercurio, Aug 2026]. The capital is earmarked for strengthening commercial strategy. The company’s stated plan is to validate its product-market fit within specific Chilean industries before expanding to the rest of Latin America, with Mexico as the next targeted market [El Mercurio, Aug 2026]. Carabela’s network of portfolio companies presents a potential beachhead of early adopters. To seed developer adoption, DocuTray offers a free tier for 100 pages per month and a startup program providing $1,000 in credits for up to 25,000 pages processed over a year [DocuTray][DocuTray, July 2026].
The Competitive and Technical Landscape
DocuTray operates in a crowded global field of document intelligence APIs, competing with established players like France’s Mindee and others such as Reducto. Its differentiation rests on a combination of factors tailored for its regional focus.
- Workflow depth. The promise isn’t just extraction, but a full, automated pipeline from scan to validated system entry. This reduces the need for custom middleware.
- Latam-specific tuning. Experience with local document formats, tax codes, and business practices from the Datalized days could provide an accuracy edge for regional clients.
- Developer experience. Offering a CLI alongside the API and a generous free tier lowers the barrier to integration and testing [DocuTray blog, Oct 2026].
The technical breakdown is straightforward. The system accepts a document, passes it through OCR and classification models, applies validation rules, and outputs JSON. The claimed sub-20-second latency is adequate for asynchronous batch processing, which covers most back-office use cases. The real engineering challenge isn’t the median case, but the long tail of document quality and format edge cases, particularly with poor-quality scans or non-standard forms.
Scaling this reliably is where the wheels could come off. The election processing was a monumental proof of throughput, but it was likely a homogeneous dataset. Enterprise sales will demand consistent accuracy across a wild variety of document types, each with its own validation logic. The 90% time-saving claim, while compelling, is also a high bar to maintain as processing complexity increases with enrichment rules and system integrations. The company’s next phase depends on moving from proving it can process documents to proving it can do so with the consistency and configurability that justifies an enterprise budget.
Sources
- [El Mercurio, Aug 2026] DocuTray levanta US$ 300 mil y se prepara para escalar a México | https://mediastation.simbiu.es/SimbiuStore/DetalleNota?detalle=Z5GLVBRDP3URV5E2QMCSPVKR7KPHZA4EUIQSRWQGEWH67VCIB7YQ&trackingToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZFVzZXIiOiIxMzEwMyIsImlkQ29ycG9yYXRlIjoiNjg2MyJ9._Kbx9qguXf0T2gdKRcoe-qOfP-A7o2Z4xcOofdPbQDk
- [DocuTray] DocuTray, API de Procesamiento de Documentos | De PDF a JSON | https://docutray.com/
- [Carabela Ventures, July 2026] Why We Invested in DocuTray | https://www.linkedin.com/pulse/why-we-invested-docutray-carabelavc-deidc
- [DocuTray, July 2026] DocuTray para Startups: USD 1.000 en créditos | https://docutray.com/startups
- [DocuTray blog, Oct 2026] DocuTray - Document Processing API | https://blog.docutray.com/
- [RocketReach] Matías Sánchez Cabrera profile |