The Open Knowledge Foundation is glad to announce today the release of the new version of the Open Data Editor (ODE), our simple, low-tech application created to detect errors in tables and enhance data literacy across non-technical communities.
With this release, ODE evolves into a key companion tool designed to support organisations in the early and critical stages of their AI journey.
The main new feature in version 1.6.0 is a new approach to the integration of artificial intelligence tools. Following intensive user research and community feedback sessions, we developed this functionality to enhance the core function of the Open Data Editor – improving the quality of uploaded data – in a non-extractive and privacy-preserving way.
A key component of this new integration is the use of responsible, offline LLMs, meaning that no data is sent to the cloud, and all operations take place on the user’s computer. This deepens our commitment to developing simple, sustainable and long-lasting technologies that solve people’s real problems, in line with our The Tech We Want initiative.
The community was eagerly awaiting this update, which will facilitate the work of the second cohort of pilot organisations – from Mongolia to Mexico – that are currently collaborating with OKFN. They are part of a growing global movement to make artificial intelligence and data literacy more accessible to everyone.
Installation
If you have the Open Data Editor’s previous versions installed on your computer, please note that the update will not be done automatically. You will need to download it again following the link below.
↪ Take the course: Learn how to use ODE
What’s New
🦾 AI feature: The ChapGPT integration was removed and replaced by a local solution powered by the llama.cpp open source tool. This new architecture will allow users to work with Large Language Models (LLMs) without sending their data to external services.
📁 Publishing feature: Based on user feedback, the Publishing feature has been replaced with a Download functionality. Users will no longer be able to download the files originally ingested into the ODE. If the data contains errors, a second version of the file will be created, highlighting the errors in cells and providing a detailed report.
🎨 UX improvements: We have made some changes to the text in the ‘Upload’ feature and removed the fixed width of the sidebar, allowing users to resize this area.
How does the new AI integration work?
- As a first step, you will need to click on the AI button, located on the upper right corner:
- Before starting to use the feature, ODE will display a dialog to inform that the AI implementation works locally:
- After clicking OK, a new dialog will pop up. Users can then select and download one of the models (sets of data) that the local implementation will use to generate prompt responses:
- Once the model has been downloaded, it needs to be selected by clicking on ‘Select model’. The tool will then display a new dialogue box containing the full list of column names in your dataset and some contextual information to ensure the model provides an accurate response:
- After clicking on ‘Execute’, ODE will display the response on the main screen. Please note that, since the AI integration works locally, depending on the number of columns in your file, this operation could take a few minutes.
Features to simplify your work
The Open Data Editor isn’t another complex data tool – it’s your shortcut to better data and improved data literacy.
Here are a few tasks that ODE 1.6.0 can help you with:
Detect errors in spreadsheets in a matter of seconds
Check if the data formats in your columns are correct
Learn data skills with an intuitive tool
Open Data Editor Use Cases
Here is how organisations across the world are using ODE:
Open Data Editor meets School of Data
In the past few months, hundreds of people around the world have received Open Data Editor training from certified instructors of the School of Data network:
All of Open Knowledge’s work with the Open Data Editor is made possible thanks to a charitable grant from the Patrick J. McGovern Foundation. Learn more about its funding programmes here.
If you have any questions or want any additional information about ODE, you can contact us at info@okfn.org.
