Privacy information
PyData Amsterdam Programme Guide · Updated 2 September 2026
Please do not enter personal, confidential, payment, health, or account information. The guide only needs questions about public PyData talks and schedules.
Who operates this guide
The guide is provided by the PyData Amsterdam organisers. For privacy questions or requests, use the organiser contact details published on the PyData Amsterdam website.
The static schedule and planner
- The read-only schedule downloads a static programme file and does not send talk selections to the API.
- When you star a talk in the planner, only its programme ID is saved in your browser's local storage on that device.
- The favorites-only view and PNG export run entirely in your browser. The image is not uploaded by this application.
- You can remove stored selections with “Clear favorites” or your browser's site-data controls.
The AI programme guide
When you submit a question, the question and limited recent conversation context are sent to the Render-hosted API. Mistral converts the question into an embedding. Chroma receives the numeric embedding and returns relevant public programme records. Mistral then receives the question together with retrieved talk titles, abstracts, descriptions, speakers, rooms, and times to generate the cited answer.
The application does not create user accounts, use advertising analytics, or intentionally save chat histories in an application database. Chat responses are marked not to be cached. Render and the other service providers may still process technical and operational information such as IP address, request time, browser type, request path, performance information, and security logs under their terms.
Purpose, legal basis, and retention
The processing is used to provide the requested programme-search service, prevent abuse, and keep the service secure. The intended legal basis is the organisers' legitimate interest in offering an optional conference guide; the non-AI static schedule remains available. The organisers should document and confirm this assessment before the public event launch.
This application does not persist submitted questions. Browser favorites remain until you clear them. Unless Zero Data Retention is enabled for the Mistral organisation, Mistral states that API inputs and outputs may be retained for 30 rolling days for abuse monitoring. Other provider operational records follow the applicable account settings, contracts, and retention policies.
Processors and international transfers
Render hosts the frontend and API; the API is configured for Frankfurt, while Render static sites use a global CDN. Mistral provides embeddings and answer generation. Chroma hosts the public talk vectors. The organisers should execute or accept the providers' data-processing terms, review subprocessors, confirm processing regions, and document any transfer mechanism such as Standard Contractual Clauses.
Squarespace embedding
When this guide is embedded on the PyData Amsterdam Squarespace site, Squarespace's own cookies, analytics, and request logs are separate from this application's data flow. The organisers should ensure the main site's cookie banner and privacy notice also cover the embedded guide and any optional Squarespace analytics.
Your choices and rights
You can use the static schedule without asking the AI guide a question. Depending on applicable law, you may request access, correction, deletion, restriction, or object to processing, and lodge a complaint with the competent data-protection authority. Contact the PyData Amsterdam organisers to exercise these rights.