Span Team
Last week we introduced Span’s AI Effectiveness Suite and the Effectiveness Scorecard, built on the new signal of agent traces and agent evals.
The next experience we’re unveiling is Spotlights: a feed of findings and best practices surfaced from analysis of agent traces across your organization.
Most engineering leaders can already feel that AI proficiency is uneven. Some developers and teams are finding better ways to work with AI. The hard part is helping more of the organization see those bright spots and build on them.
Today, those lessons are shared manually — a tip in Slack, a presentation at a lunch and learn — and they are easy to miss. Span helps surface the findings worth sharing, using agent traces and evals to identify workflows, correlations, and emerging practices that lead to stronger outcomes.
Actionable findings from across the organization
Spotlights gives developers a feed-like experience they can come back to regularly to see what findings Span is surfacing from across the organization. They can upvote the ones that seem most useful and drill in for more detail about the pattern and how to apply it to their own work.
In some cases, Spotlights helps spread something people are already doing well. In others, it uncovers findings that are not yet obvious to any one person or team, helping the whole organization learn faster.
By default, these Spotlights are not attributed to individuals. The focus is on making useful, data-backed ideas easier to discover and adopt. Over time, this also creates a foundation for highlighting the people and practices that are having outsized impact.
What leaders see
When stronger workflows and practices become easier to see and adopt, they spread more naturally across teams. As more people move up the proficiency curve, organizations get more from the tools and spend they already have in place.
More soon on what comes beyond individual proficiency, and why the environment matters too.