Extend’s Theranos Desk Simulator Uses Real Court Documents

Extend, a B2B SaaS firm that claims to parse, extract, and split documents with unmatched accuracy, released a web application that lets users sit at Elizabeth Holmes’ desk in 2016. The site pulls over a thousand emails, slides, texts, and machine logs that were disclosed during the United States v. Holmes criminal trial and renders them on a virtual MacBook Air running OS X El Capitan and an iPhone with a home button.

The environment is interactive. Users can click on the laptop or phone screens to view the original PDFs, scroll through the documents, and even launch a simulation of the Theranos Edison device, the company’s blood-testing machine that promised to extract large amounts of health data from a single drop of blood. The simulator is built with standard web technologies; the underlying data is served from Extend’s own document-processing pipeline, which the company says can handle complex, unstructured files with high precision.

Bo Lau, an engineer at Extend, posted the project on Twitter on October 8, 2026, noting that the site was created as a way to make the court exhibits more accessible. The project also doubles as a promotional tool: Lau is hosting a watch party for Nathan Fielder’s upcoming documentary, “You Can See Everything,” and is advertising Extend’s product alongside the event.

How the Simulator Works

The simulator’s core is a rendering layer that maps court-exposed files to a 3-D office scene. The files themselves are stored in Extend’s document-parsing service, which tokenizes PDFs, extracts text, and indexes metadata. When a user clicks on an icon, the service streams the original file to the browser, preserving formatting and annotations. The Edison simulation is a lightweight JavaScript model that mimics the device’s user interface and displays sample output data.

Extend’s marketing claims that its parsing engine can handle “hardest documents with unmatched accuracy.” In practice, the simulator demonstrates the engine’s ability to ingest large, multi-page PDFs, preserve hyperlinks, and expose searchable text. The service also splits documents into logical sections, a feature that could aid legal teams in quickly locating relevant passages.

Developer Implications

For developers, the simulator offers a concrete use case for document-parsing APIs. The ability to ingest court exhibits, preserve structure, and expose interactive previews could be valuable for e-discovery platforms, compliance tools, and knowledge-management systems. Extend’s open-source SDKs expose endpoints for text extraction, metadata retrieval, and document segmentation, allowing teams to integrate similar functionality into their own workflows.

The project also highlights the importance of data provenance. By sourcing files directly from court exhibits, Extend demonstrates compliance with legal disclosure requirements and offers a transparent audit trail. Developers building compliance-focused products can use the simulator as a benchmark for accuracy and traceability.

Marketing and Monetization

Extend is positioning the simulator as both a demo and a marketing vehicle. The company’s website includes a call-to-action for potential customers to schedule a demo of the parsing platform. The watch party for Fielder’s documentary serves as a secondary promotion, leveraging the public’s interest in the Theranos saga to drive traffic to the site.

While the simulator is free to use, Extend’s business model remains subscription-based. The company offers tiered plans that scale with document volume and feature set, including advanced analytics and integration hooks. The simulator’s public availability may lower the barrier to entry for new customers, but the true value lies in the underlying API that powers the interactive experience.

Risks and Caveats

The simulator relies on court-exposed data that is subject to legal restrictions. Extend’s use of the data is permissible under the public domain status of court exhibits, but developers must ensure compliance with any post-trial confidentiality orders. The accuracy of the parsing engine is demonstrated on a limited set of documents; performance on other file types or languages remains untested.

The project also raises questions about the ethical use of sensitive medical data. While the Edison machine’s output is simulated, the original documents contain patient information that was disclosed in court filings. Extend’s handling of such data must adhere to HIPAA and other privacy regulations, a factor that developers should consider when integrating the API into healthcare workflows.

What to Watch Next

Extend’s public demo signals a broader trend of companies using real legal data to showcase document-processing capabilities. Future releases may include additional use cases, such as automated redaction or contract analytics. Developers should monitor Extend’s roadmap for API updates that expand language support and integration options.

The Theranos desk simulator also serves as a case study in how to turn a high-profile legal event into a product showcase. The combination of interactive visualization, real data, and a marketing hook could become a template for other SaaS firms looking to demonstrate their technology in a tangible way.

Related Coverage

The simulator’s launch coincides with the upcoming TechCrunch Disrupt 2026 event, where Extend plans to present its parsing platform. For more on the event’s AI focus, see the TechCrunch Disrupt 2026 Offers $200 Ticket Discount as AI Leaders Converge.

Related coverage

Explore more on this topic