What's new on LaunchDarkly

Context counts chart

If you’ve ever wondered “how many users have seen my new feature?” you can now use the new Context Counts chart to answer that question! On the Monitoring tab of any flag that has client-side evaluations, in addition to the Evaluations chart, you can now also select a context kind to view the number of unique contexts of that context kind that encountered the flag.

For now, this is only supported for flags with client-side evaluations. Learn more in the LaunchDarkly docs.

AI Configs: Agents, Tools, Trends

Shipping AI-powered features and agent-based workflows is the new normal. It’s no longer a question of whether you ship them, but how you do it safely, consistently, and with control. We have just introduced three powerful updates to AI Configs:, agent-based workflows, tools, and trends explorer for AI Insights.

Agent-based workflows

Now, with AI Configs, teams can define agent behaviors, attach reusable tools from a shared library, and use new SDK methods to spin up intelligent, multi-step agents, while still benefiting from the safe rollout and approval flows you expect from LaunchDarkly.

Tools

For agent-based workflows to be effective you also need effective tools. These can be used with both our Agent and Completion modes, providing ultimate flexibility in how LLM workflows are set up. Tools allow LLMs to do things like retrieve extra data or take an action on a users' behalf. Tools can now be managed in the UI using either a visual or JSON editor to define the configuration, are versioned on update so you have a change-log of any edits made to your tools, and can be attached to either Completion or Agent configs.

Trends Explorer for AI insights

As your team ships more AI products, it becomes harder to measure the real impact of model behavior, performance, and cost across environments and teams. With the new Trends Explorer, you can visualize trends across your AI Configs. Quickly visualize which models cost the most, have the greatest latency, or are performing the best (or worst).

Get started with AI Configs today. AI Configs Docs | Start your free trial

Data Export for BigQuery and Databricks

Data Export allows users to export their flag evaluation events, experiment / flag metadata, and metric events to their own data warehouse. We know that Enterprise customers with an existing data warehouse strategy often want to overlay their existing product and business reports with flag data to answer more detailed questions about their users and products. Now, customers using Databricks and BigQuery who are interested in flag data have Data Export support.

See our documentation for setup instructions: Databricks Data Export | BigQuery Data Export

Other improvements

• Released LaunchDarkly for Jira (EU) - find it on the Atlassian Marketplace
• Made default contexts available for guarded rollouts and experimentation
• Made client-side flag evaluations visible directly in the session player timeline
• Added SDK support for Node.js, Python, and React Native in LaunchDarkly Observability
• Enhanced Observability Trace Viewer to provide full context in a single view
• Enhanced data visualization for metric details when chart is hovered
• Improved accuracy of cost calculation logic for AI Configs
• Improved variation value filtering in events API/service layers