From Standups to Sprints: Adopting Agile Ceremonies for AI in Social Impact Work

This is the third of a series on Agile Management and AI in the Social Impact Sector.

In the first two blogs in this series, we covered some of the concepts and organizational culture shifts necessary to drive agile processes.  This blog encourages us to go a level deeper by adopting the specific meetings and group processes that make up agile.

As reminder as to why agile is needed, implementing AI within a social impact organization often feels like building the plane while flying it. The fast pace of AI innovation combined with complex ethical and human-centered considerations can be overwhelming. That’s where the Agile "ceremonies" that make up agile project management come in—not as a rigid process, but as a rhythm to guide experimentation, learning, and responsible progress.

What Are Agile Ceremonies?

Agile ceremonies are recurring meetings that help teams align, plan, and reflect on their work. While rooted in software development, they’re increasingly valuable in mission-driven environments where collaboration, iteration, and transparency are essential. 

Here’s how social impact organizations can adopt these ceremonies to drive quickly-changing, meaningful and ethical AI work:

1. Daily Standups: Short Check-ins, Big Clarity

What is it?: A 15-minute meeting where each team member shares what they did yesterday, what they're doing today, and any blockers. 

Structure:

• Duration: 15 minutes

• Format: Each project team member answers the following questions.

  • What did I work on yesterday?

  • What am I working on today?

  • What blockers am I facing?

In practice for AI work:
Standups are great for aligning data scientists, program staff, and policy teams working on an AI solution. The identification of blockers should not delve into a discussion around solves. The team leader should address blockers after the meeting to unblock individuals.

Tip: Use this time to make ethical concerns just as valid as technical blockers.

2. Sprint Planning: Setting Clear Goals for Short Bursts

What it is: A meeting at the start of a 1-2 week sprint to define what work will be completed.

Structure:

• Duration: 60–90 minutes

• Inputs: Backlog items ( a list of tasks maintained by the team leader, e.g., model accuracy targets, API integrations, stakeholder feedback loops)

• Outputs: A prioritized task-level list with time to completion estimates that fit within the 1 or 2 week sprint work time. 

In practice for AI work:
During sprint planning, teams can prioritize tasks like curating training data, refining an AI model, or testing output fairness. By limiting the work duration and creating the opportunity to reprioritize, teams can manage new information such as client feedback and technological changes. 

Tip: Mix technical and domain expertise when defining sprint priorities to keep impact top-of-mind.

3. Sprint Review: Demo What You Built, Hear What They Think

What it is: A show-and-tell session at the end of a sprint where teams demo their work to stakeholders.

Structure:

• Duration: 45–60 minutes

• Attendees: Cross-functional team + stakeholder reps

• Focus: Live demo of AI model, dashboard, or user-facing tool + feedback loop

In practice for AI work:
The demo gives team members to show off their accomplishments while creating the opportunity for stakeholders to provide input at each step in the process of AI development as opposed to waiting until the end of a project. Feedback can be priortized and immediately incorporated into the next sprint.  For example, imagine a social enterprise developing an AI model to detect food insecurity patterns. A sprint review might include a demo of the dashboard and an open conversation with community partners to evaluate if the data insights are actionable—and respectful.

Tip: Invite external stakeholders—like community members or ethicists—to your reviews for diverse feedback.

4. Sprint Retrospective: Reflect, Learn, Improve

What it is: A time for the team to reflect on the past sprint—what worked, what didn’t, and what could be improved.

Format: Hold a 30-45 minute of project team members focusing on the following questions.  

• What went well?

• What didn’t go well?

• What should we try next?

In practice for AI work:
This is the moment to ask: Were all voices heard in the design process? Did the AI behave unexpectedly? Were ethical guardrails effective? For organizations piloting an AI recruitment tool, this could surface crucial concerns around bias or accessibility.

Tip: Don’t just focus on the tech—make space for team dynamics, values alignment, and mission impact.  

Getting Started

Here’s how a social impact team new to Agile can ease into ceremonies:

• Start small: Try just standups and retros for the first few weeks.

• Use visuals: Kanban boards or sprint management in project management system, or shared docs help track sprint goals.

• Embrace learning: Every ceremony is a chance to listen, learn, and adjust.

Final Thoughts

Agile ceremonies aren’t just for coders—they’re for mission-driven teams navigating complexity, change, and innovation. When applied thoughtfully, these rituals help social impact organizations bring transparency, adaptability, and shared purpose to the evolving world of AI.

By grounding your AI journey in these rhythms, you're not just building technology—you’re building trust, accountability, and social good.