Proposal Guidelines

You may submit up to 2 talk proposals this year. Please do not submit more than 2 proposals. If you are unsure whether your talk is a good fit, we encourage you to submit and let the Program Committee decide.

For each talk, PyData Amsterdam can offer free tickets for up to 2 speakers. If there are more than 2 speakers, the additional speakers will need to buy a ticket separately.

The proposal selection process is double-blind, meaning that reviewers will not be able to see your name or other identifying information. Please withhold identifying information (such as a personal photo) from those fields. You can add this later if your proposal is accepted.

Talk Formats

Standard Talks : 30 minutes total (including 5 minutes Q&A)

Deep Dive Talks* : 45 minutes total (including 5 minutes Q&A)

*Limited availability. You may indicate interest during submission, but allocation is not guaranteed.

Proposal Structure

Each submission consists of two required components.

Abstract (Public Description)

The abstract explains your talk to potential attendees and will appear in the official conference schedule.

A strong abstract should:

• Summarize the topic and its importance

• Identify the intended audience

• Highlight key takeaways

• Clearly communicate what attendees will learn

• Remain concise and engaging

Think of the abstract as a short pitch motivating attendees to join your session.

If supporting materials will be shared (e.g., notebooks, datasets, repositories), specify how participants will access them.

Description (Program Committee Review)

The description expands upon the abstract and provides sufficient detail for the Program Committee to evaluate:
- Technical depth
- Scope feasibility
- Educational value
- Relevance to the conference

Including an outline with estimated timing is strongly encouraged, especially for talks covering multiple topics. Anticipate potential reviewer questions and address them proactively.

What Makes a Strong Proposal

A proposal should convince attendees to spend 30–45 minutes learning from you. Strong proposals clearly communicate the following:

• Topic and Relevance: What will you present, and why does it matter now? Explain the insight, practical value, or lessons attendees will gain.

• Clear Title: Your title should accurately reflect the content of your talk. Creativity is welcome, but clarity is essential.

• Target Audience: Specify who will benefit from attending, including relevant roles (e.g., data scientists, engineers, researchers) and experience level.

• Audience Takeaways: Describe the concrete knowledge, skills, or perspectives participants will leave with.

• Talk Type and Approach: Indicate how the material will be presented (e.g., conceptual, mathematical, practical, demo-driven, hands-on).

• Required Background Knowledge: List any tools, concepts, or prior experience attendees should have to follow the talk.

• Scope and Structure: Provide a realistic overview of how the talk will be organized. A high-level timing breakdown is encouraged.

Example:
- Introduction — 5 min
- Problem Context — 10 min
- Case Study or Demo — 10 min
- Lessons Learned and Q&A — 5 min

Writing Tips

A strong proposal serves three purposes:
- Enables the Program Committee to confidently evaluate your talk
- Becomes the public session description
- Helps create a diverse and balanced conference program

Recommendations:
- Ask colleagues, especially your target audience, for feedback
- Clearly define scope boundaries (for example, “We focus on X and will not cover Y”)
- Keep proposals concise and focused (around 200 words is recommended, but not required)

If accepted, you will have an opportunity to refine wording before publication. Edits must remain within the original scope of submission. Additional speakers cannot be added after submission.

Use of AI Tools

Using AI tools to improve writing quality is acceptable when done responsibly.

Appropriate uses include:
- Improving clarity, structure, or grammar
- Translating proposals into English
- Checking tone or readability

Please avoid:
- Submitting AI-generated proposals you cannot confidently deliver
- Inventing results, benchmarks, or case studies
- Claiming experience you do not have
- Submitting mass-produced or near-identical proposals

A Call for Proposals is not a call for prompt engineering. Reviewers value authentic expertise and clearly articulated experience. 

Review Process

Submissions are reviewed using a double-blind process. Reviewers will only see:
- Title
- Prior Knowledge Expected
- Brief Summary / Abstract
- Bullet Point Outline
- Description
- Specialized Tracks
- Keywords

Please do not include identifying information in these fields, including:
- Names or affiliations
- Company or organization names
- Personal websites or profiles
- Projects uniquely tied to you

Common Pitfalls

Proposals may be declined when they are:
- Previously recorded talks already available online
- Overly long or unfocused
- Primarily based on unfinished future work
- Sales or product pitches
- Generic or fully AI-generated descriptions lacking technical depth

PyData prioritizes educational, practical, and community-focused content centered around the open-source ecosystem.

Speaker Benefits

Accepted speakers receive a free full-conference ticket.

If exceptional circumstances arise (such as illness, visa issues, or unavoidable conflicts), please contact the organizing committee at: amsterdam@pydata.org

Speaker Mentorship Program

If this is your first time presenting to a live audience, you may request participation in the PyData Speaker Mentorship Program.

Participants are paired with experienced community speakers to help prepare and refine their presentations. Please indicate interest in the submission form.

FAQ

Do speakers receive free conference access?
Yes. Accepted speakers receive a full conference pass.

My company wants to sponsor. Can I introduce them?
Yes. Please indicate this in your application. Sponsorship interest has no influence on proposal acceptance but helps support open-source and educational outreach through NumFOCUS.

My proposal is not about LLMs or current AI trends. Should I still submit?
Yes. PyData represents the full breadth of the data science ecosystem, including statistical modeling, data engineering, visualization, scientific computing, time series analysis, optimization, geospatial analysis, and many other topics. Non-trend topics are strongly encouraged.

Submissions close on 2026-04-01 23:59 (Europe/Amsterdam).