Call for papers

Intro

This is your invitation to submit a talk for Local-First Conf 2026.

We’re open to anyone working in computing technology (products or research) that shares our values of data ownership and user empowerment.

Submit your first low-effort application (a few questions, less than five minutes of your time) in our initial phase. Finalists will then be invited to flesh out their talk description before the final selection.

Timeline

Feb 26

CFP opens

MAY 1

CFP closes

MAY 15

Candidate speakers notified

June 1

Speakers announced

June 15

Schedule announced

Call for topics

We welcome talks on any topic you think is relevant to user empowerment and data ownership.

However, here are some topics we're especially interested in:

New Territories

We're calling for everyone building better tooling for local-first architectures. What projects under active research and development are pushing the envelope of what user empowerment means? We want to hear from builders exploring new frontiers and experimenting with emerging technologies that put users in control.

Design for agency

As we see new AI technologies creating more expansive—and more reckless—uses of user data, what design patterns can afford users more control over their information? Where is local-first well-positioned when it comes to granting AI agents capabilities while keeping users in control? How do we design interfaces that help users understand what's happening with their data and make meaningful choices about how LLMs interact with it?

We're looking for research projects and real-world products that give users control and confidence when it comes to LLM interactions. Show us your work on local-first designs tailored to AI usage, patterns for transparent agent behavior, or interfaces that help users maintain agency even when working with powerful AI systems.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

Local and open weight models

Running AI models locally on users' own devices opens up new possibilities for privacy, offline capability, and true data ownership. But training, optimizing, and distributing models that can actually run well on consumer hardware is a significant challenge. What does it take to make powerful AI accessible without requiring users to send their data to the cloud?

We'd like to hear from someone who has done significant work on training AI models or building the supporting infrastructure that enables users to install and run models on their own devices. This could include work on model families like Llama and Qwen, infrastructure like Ollama or PyTorch, or optimization techniques that make local inference practical for everyday users.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

AI assistance for local-first

AI is dramatically lowering the cost of integrating with third-party services and breaking down data silos. Where users once depended on official integrations between apps (or gave up and lived in silos), AI agents can now build custom connectors, exporters, and sync bridges on the fly—giving users real ownership over their data even when the original app doesn't make it easy. Beyond integration, AI opens the door to cross-app interoperability without requiring apps to agree on a shared data model through AI-generated mappings and dynamic lenses that adapt in real time.

We'd like to hear from anyone building at this intersection: AI-assisted data migration, dynamic schema translation, agent-driven integrations, or other ways AI is helping users break free from cloud silos. This picks up where projects like Ink & Switch's Cambria left off, but with LLMs as the translation engine. Show us your practical demos, real-world usage, and vision for how AI can empower users to truly own and control their data across any application.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Local-First Maturity

This is the third installment of Local-First Conf—what emerging trends matter most to local-first builders? We want to hear your real-world stories, the hard lessons you've learned, and how you've solved your 'last-mile' problems. Share your experiences building production systems that put users first.

Local-first success stories

Do you run a commercial or open-source product that has shown real scale with syncing, CRDTs, self-sovereign identity, e2ee, or other local-first technologies? We want to hear your "report from the field" on how local-first has worked in practice with real users and real data. What does it actually look like to ship and maintain a production local-first system?

A strong talk will include metrics (number of users, quantity of data, scale challenges you've faced) and your honest war stories: what worked well? What was a catastrophic failure and had to be rewritten? What are the hard parts that surprised you? Share the genuine differences you've discovered compared to traditional architectures (pure local or pure cloud), and help the next generation of builders learn from your experience.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

UX challenges of local-first

Classic software had a well-understood pattern: files, save, load. The cloud also brought a universal pattern: you have an account, and your stuff is "inside" your account. But local-first is still a developing space with emerging UX patterns that users are still learning to understand. Where is my stuff? How do I search, back it up, or share it with another person? How do I revoke share access? How do I manage multiple versions when things are out of sync? These are the hard questions we're still figuring out.

We'd like to see submissions from people with significant experience working on local-first research or products with real-world users. Draw from those experiences to give us concrete examples of what works and what doesn't. Share the design patterns you've discovered, the mental models that clicked with users, and the challenges you're still wrestling with.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Legal frameworks

Data ownership and user empowerment aren't just technical challenges, they're legal ones, too. How do GDPR and other regulations fare when it comes to putting power and control back in users' hands? What are the liabilities and threats that come with data ownership? As local-first and edge computing reshape where data lives and who controls it, we need to understand the legal landscape.

We want to hear from lawyers and policy makers who have worked in this field and evaluated different approaches. Share your experience navigating privacy regulations, data sovereignty laws, and the legal frameworks that either enable or constrain local-first architectures. Help us understand the legal realities, as well as upcoming opportunities, for building systems that truly give users control.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

The Larger Ecosystem

"Local-first" is a term made popular by a 2019 essay, but the idea of user empowerment and data ownership is older and broader. We want to connect with people who have built systems in adjacent communities so we can grow together. If you're working on technology that gives users more agency over their data and digital lives, we want to learn from you.

Self-sovereign identity

Most ways we identify ourselves to computer systems involve a central registry controlled by an institutional entity. Local-first systems can use cloud-style logins, but we'd rather have something that grants more ownership over that identity to the individual. The challenge is creating identity systems that are both secure and truly user-controlled.

There have been many attempts over the years (OpenID, Microsoft Passport, Mozilla Persona) that haven't gained widespread adoption. If you've worked on one of these systems or have significant experience with them, we'd love to hear your war stories. Or if you're working on an identity system today, tell us how you're solving this problem and how you'll avoid the challenges that blocked past attempts from succeeding.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Social media and data ownership

The OGs of social media—Facebook, Twitter, Instagram—are the exact opposite of the user agency and empowerment we want to see in the world. Data silos that are hard to access, manipulative algorithms that decide what users see, and limited programmability or data export are the norm. But new protocols and networks are trying to change this, with Mastodon's ActivityPub and Bluesky's ATproto leading the way. How are these communities giving users real ownership of their content, connections, and social graphs?

If you're deep in building one of these communities, we'd like to hear a state-of-the-world report from you. Give us an overview of the user-facing app or protocol (including usage metrics), then tell us how it's working and how your community is bringing user agency and empowerment back to social media. Share concrete examples of what users can now do that they couldn't before, and give the audience a call to action for how they can get involved and be part of your movement.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Encrypted messaging

End-to-end encrypted messaging apps (such as Signal, WhatsApp, and Matrix) have established a standard for secure communication and offline capability for users. At the same time, they've faced many UX challenges compared to their non-e2ee competition—for example, accessing chat history on a new device or managing identity across multiple devices. These apps have been solving local-first problems for years, often before we had a name for it.

We'd love to hear talk submissions from anyone who has worked on one of these apps or other real-world applications of e2ee, zero-knowledge systems, or privacy-preserving technologies like Tor. Share the hard lessons you've learned about balancing security, usability, and performance. Tell us about the trade-offs you've made and the design patterns that have worked (or failed) when building systems that must be both secure and user-friendly.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Game development with local-first

Video games have a long tradition of requiring extremely responsive interfaces (60fps+) and handling complex state synchronization. These are challenges that overlap deeply with local-first principles. Game developers have been solving problems around offline play, peer-to-peer multiplayer, local mods, and user-generated content for decades. What can the local-first community learn from how games handle data persistence, conflict resolution, and state synchronization?

We'd love to hear from someone who has worked on building games directly (for game consoles, mobile, PC, etc.) or game development tools (Unity, Godot, Unreal, custom level editors, etc.). Share how you've approached persisting and sharing game data, handling multiplayer state, enabling mods and customization, or any other aspect of game development that gives players more control and agency over their gaming experience.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Malleable software in the age of LLMs

At Local-First Conf 2024, Maggie Appleton's "Barefoot Developers" painted a picture of a world where data ownership and more accessible programming could add up to a new class of developers. Now in 2026, LLM-assisted coding may have unlocked new doors for barefoot developers and for malleable software—systems that can be modified by their users. What happens when end users can ask an AI to reshape their tools to fit their needs? How do we build software that's designed to be changed?

We'd like to hear your submission if you are a barefoot developer who has used AI to modify software for your own needs, or if you've worked on a real-world product that enables greater software malleability. Show us examples of users customizing their tools, share the technical architecture that makes modification safe and practical, and tell us about the new possibilities that emerge when software becomes truly malleable.

Apply to speak

Suggest someone

Share this with someone who you
think should speak on this topic!

Copy to clipboard

Apply to speak

Suggest someone

copy to clipboard