Report from the Jupyter and AI community meetup

Thu, 23 Apr 2026 00:00:00 +0000

A small group of Jupyter community members recently met in Seattle to demo Jupyter workflows augmented with AI tools and discuss what’s next. This is a quick report-out about what stood out.

The conversations focused on what agentic workflows could unlock for researchers and students - with broad acknowledgement that deployment, security, authorization, and ethics still have a lot to be worked out.

Here are a few highlights from the demos.

Jupyter AI and the Agent Client Protocol #

Jupyter AI v3 is adding support for the Agent Client Protocol (ACP), enabling multi-agent collaboration within JupyterLab. We saw how a single request could be split across multiple specialized agents based on each one’s skills, similar to this demo from JupyterCon 2025.

Some things still to figure out:

No tool isolation yet - agents aren’t scoped by user role or security group.
Manual configuration required - agents running locally need to be registered by hand.
A future goal is for JupyterLab to auto-detect locally installed agents, or for LLM providers to contribute integrations directly.

The Jupyter AI team also shared a new “Personas” abstraction that lets you define a skill-specific agent in just a few lines of Python. Personas open the door to domain-specific agents or agents with crafted “personalities”. They are written in a few lines of Python, though we discussed the possibility of using markdown instead. This seems to be evolving rapidly!

Jupyter AI v3.0.0 has since been released.

Notebook CLI #

nb-cli is a Rust-based tool for interacting with notebooks from the command line - changing cells, running them, checking errors, and rendering the notebook in the console. It can also connect to remote JupyterLab instances, making it useful for headless workflows like CI pipelines or batch grading.

Sidebar comments and Real-time collaboration #

A demo showed Google Docs-style side-panel comments for .ipynb files. Previous collaboration efforts used Conflict-free Replicated Data Types (CRDTs) for multi-cursor editing, but the team felt that approach may not have hit the mark. Side-panel comments felt more natural, though how this extends to .py files is still an open question.

Opportunities for 2i2c #

The meeting helped us understand the role we could play in this space. We think there’s opportunity to help facilitate the use of agents on shared infrastructure, particularly around the upstream control planes for deployment, administration, and authorization. We can also help facilitate community conversations about the ethics and impact of AI tools on open science.

Acknowledgements #

The Jupyter Foundation for supporting the event.
Zach Sailer (Apple), for organizing this meetup.
Chris Holdgraf for editing and adapting the original content for this post.

Improvements to our team's planning and delivery

Mon, 15 Apr 2024 00:00:00 +0000

This is a follow-up to our 2023 report of organizational strengths and weaknesses, describing some improvements we’ve made on our team’s coordination and delivery.

In 2023, we released a report describing our organizational strengths and weaknesses. This uncovered a key challenge for our team: improving our coordination and delivery. Over the previous two years, our service had grown significantly in its scope and complexity. We were working on more than 7 active grants and were serving more than 70 active communities with around 7,000 monthly active users.

This was taxing our small team, and we found ourselves struggling to efficiently deliver on our work. For example, a collaboration with GESIS to bring image building functionality to JupyterHub took longer than we wished, and we felt that our planning and execution was not transparent enough to their team. In addition, in a collaboration to serve communities in Latin America and Africa we felt that 2i2c was not responsive enough to onboarding and deploying infrastructure for new communities.

Steps we’ve taken to improve delivery #

At the beginning of 2024, we hired our first Delivery Manager and Chief of Staff role as well as our first Product Lead role. Both of these roles are meant to develop systems and team practices that improve our planning and delivery.

In Quarter 1 of 2024, we designed and initiated an organizational transformation with the following goals:

Provide clarity about our overall organizational goals and strategy.
Define our near-term goals and major projects that drive our work.
Define and prioritize the major work items that feed into these goals.
Break this major work items into actionable items that our team can work on from day to day.
Provide visibility for all of this information across the entire organization.

Below is a brief description of the major changes that we’ve made. These are still a work in progress, and organizational transformation takes months, if not years, but we hope that this provides a useful snapshot in time as we kick off this process.

Our system of work #

Our team-wide system of work can be found in the Team Compass at the link below:

compass.2i2c.org/cross-functional/workflow/

This system of work attempts to link our strategic goals with concrete chunks of work to deliver. You can see an overview of this system below:

Our work system shows how initiatives are made up of lists of actions that accomplish them. These actions are distributed across our team’s operational boards for delivery.

It is roughly broken down into these major areas:

Our value proposition: Our system of delivery starts with a value proposition. This is a north-star statement for the value that 2i2c aims to provide to our communities in order to achieve our mission. It is a guiding principle for where we prioritize our time and improve our service.

While updating our delivery model, we decided it was time to update the value proposition we’d been informally using. We’re in the process of validating this value proposition with communities, and will share a draft for public comment soon!

Strategic goals and major projects: Next we define strategic goals that describe the most important progress we must make as an organization. This considers our current capabilities and challenges, as well as the major projects that we’re already committed to (like grants). Our initiatives (described below) should each represent progress towards one or more of our goals and major projects.

Strategic initiatives: These are major thrusts of work that represent progress towards our goals. They range in time from weeks to months, and generally require coordination and action across each of 2i2c’s functional areas. Initiatives exist in a dedicated board, where we shape and scope them with enough information to understand them and prioritize. Once an initiative is in progress, it begins driving work on a two-week planning cycle. All in-progress initiatives should make up a significant percentage of 2i2c’s total work allocation.

Operational boards: Operational boards are used to track our day-to-day workstreams. We organize into two-week sprints, with a collection of work pulled into each sprint according to in-progress initiatives and the other types of reactive and operational work on our plate (for example, responding to support tickets is reactive work). Within each initiative, we coordinate across our teams in order to understand the next steps needed and who is responsible for doing it. This helps teams plan the work for their next iteration.

This system is very much still a work in progress, and we’ve already identified a number of ways that we’d like to improve it moving forward. For example, we’d like to find more efficient ways of coordinating across our team, and encouraging team members to own their own work.

Changes to our team culture #

In addition to our system around planning work, we’ve also identified a number of ways that we can improve our team’s culture related to work. As a distributed organization, one of our biggest challenges has been the additional friction that comes with communicating across many time zones. It’s much harder to solve a problem or get help when responses come with long gaps of time.

As a result, we’ve defined several practices that will help our team members grow their autonomy and independence, while still moving towards the same targets and goals. This is ongoing work that we aim to continue developing over the next few months.

Professional growth. First, we’ve started an audit of 2i2c’s organizational roles and the responsibilities and the pathways for growth with each of them. This began with our engineering team, but will continue with other areas of the organization as well. We believe that defining professional growth trajectories will help identify the gaps in our team’s skills that make it more difficult to act efficiently and independently.

Giving and receiving feedback. We’ve defined team practices around giving feedback, and are encouraging a team culture of regular and constructive feedback to one another. We know that this will be a long-term investment in team practices, and we believe that any team (especially a distributed one) must be able to rely on its team members to help one another grow. We’ve begun this effort by creating a guide to giving feedback for 2i2c which we hope will be useful for others as well.

Decision making. We’ve defined a set of principles around decision making at 2i2c that encourage more autonomy and creativity. In a knowledge-driven field such as ours, it is important that team members have the freedom to be creative and take risks, and that we design systems that are resilient to mistakes. We aim to make mistakes Safe to Fail to encourage creativity and learning while de-risking major negative consequences. This document is a first step towards improving our team-wide practices at taking more initiative and action.

Iterative delivery and process improvement. Across each level of the organisation, we have been experimenting more deliberately with Agile delivery practices. These experiments shape our approach to planning (at both the team and organization-level), visualizing and managing our work streams, and improving our internal product delivery processes. We are agnostic to a specific Agile framework, and have found the most value in blending elements of Kanban, Scrum and Flight levels. We will continue to experiment with these practices, seeking to find a happy medium for 2i2c. We’ll share more on this soon :)

Coaching. Some members of our team have also begun benefiting from having regular 1:1 coaching to help their interpersonal development. Our team’s coaching support includes work to empower improving self-awareness, overcoming self-limiting challenges and gaining technical leadership mastery. Coupled with our growing practices around team feedback, we aim to build a culture where team members have both internal and external support to help them grow.

This work was done in partnership with

April Johnson, a Transformation Consultant and Coach. If you’re interested in working with April around any of this work we highly recommend doing so!

A continuous process of improvement #

We are still in the early days of this transformation. We hope that these practices will improve both our operational efficiency and also create a more enjoyable work environment where team members are empowered to have impact in the ways that they see best. We’re excited to lean into these challenges and grow as a team, because this is how we can grow the impact of 2i2c and deliver more value for the communities that we serve.

Harold Campbell

Mon, 01 Jan 0001 00:00:00 +0000

I enjoy building and supporting communities of practice. I have built a small JS visualization library and dabbled in delivery management systems. Recently, I have strong interests in music (programming VST, etc.), AI and data science.

Highlights:

15+ years industry experience spanning companies in Africa and Jamaica
10+ years experience in agile consulting and coaching in technical and product teams
Core skills span Enterprise Agile Coaching, Change Management, DevOps, and cloud consultations.

Harold Campbell | 2i2c