Image to Excel (formerly It Excel) vs Matrices

Upload a JPG or PNG and Image to Excel rebuilds the table rows and text inside as a downloadable .xlsx file. The web converter is free, needs no account, and returns results in a few seconds after you click Convert.

Unlike plain OCR tools that dump unstructured text, Image to Excel targets table grids and form layouts in photos. It handles English, Simplified Chinese, and Traditional Chinese, which covers a lot of scanned receipts, invoices, and paperwork that would otherwise need manual retyping.

You can use it in a browser or through dedicated iOS and Android apps. The interface is built around a single upload-and-convert flow, so the whole process takes a few clicks.

It is aimed at accountants, data entry staff, office admins, and anyone who regularly receives table data as screenshots or scanned images instead of native spreadsheet files.

Matrices transforms data analysis by combining AI-powered spreadsheets with an innovative platform for training AI agents in simulated environments. It automates research and data gathering by using AI agents that fill spreadsheets with data from multiple sources, providing confidence scores for transparency. Beyond spreadsheets, Matrices is pioneering a system of virtual 'escape rooms'—complex simulations where AI agents learn tasks through trial and error without real-world risks. This approach enables AI to practice and improve on tasks like market research, coding, and business operations in safe, scalable environments.

The platform supports two main user roles: engineers who build configurable simulation components, and creative designers who craft scenarios for AI to solve. This ecosystem aims to scale AI training by crowdsourcing simulation creation, eventually covering a wide range of real-world tasks. Matrices targets founders, investors, sales professionals, CEOs, and AI researchers who benefit from automated data workflows and advanced AI training environments.

Matrices emphasizes natural language commands to reduce the learning curve, allowing users to automate complex workflows without coding. Its extensible design adapts to diverse needs, from simple data tasks to intricate AI training challenges. The company is actively expanding its simulation library and plans to open its level editor to the public, inviting collaboration to accelerate AI automation across industries.

Technically, Matrices leverages reinforcement learning by placing AI agents in simulated environments where they can experiment and learn from feedback. This method moves beyond imitation learning, enabling AI to achieve higher competence in practical tasks. The platform’s modular architecture supports integration with various web apps and data sources, making it a versatile tool for both data analysis and AI development.

Overall, Matrices offers a unique blend of AI spreadsheet automation and a scalable framework for training AI agents in realistic, risk-free settings. This dual focus positions it as a forward-looking solution for businesses and AI practitioners aiming to automate complex workflows and accelerate AI capabilities.