Latest AI and tech news

Coding agents are quickly becoming the most common way people interact with LLMs. TruLens has spent years building evaluation and observability for agents — now we're bringing that to coding agents. This release ships client hooks for Cursor, Claude Code, and OpenCode, with more clients planned. Ins

A response can look good by itself while the conversation around it fails. TruLens 2.12 adds first-class conversation tracking and evaluation so you can measure context retention, consistency, and progress across turns without losing per-turn metrics and traces.

An LLM judge can produce a number for every response and still be a poor evaluator. It may rank examples in roughly the right order while consistently scoring too high, or look accurate on average while making costly mistakes near a decision threshold. Before relying on a judge, we need to compare i

TruLens 2.8: the Run API now works on any backend with parallel execution (up to 5.4x speedup), there's a newSchemaValidatorfor programmatic output checks, and the dashboard leaderboard is 2-5x faster via SQL aggregation. Database session initialization drops from 1.6s to 0.4s without uninformative

TruLens 2.7 brings a cleaner evaluation API, first-class MLflow integration, improved Snowflake support, and a growing library of examples—making it easier than ever to evaluate, iterate, and trust your AI applications.

Building and evaluating LLM applications just got a whole lot easier. TruLens 2.6 brings powerful new capabilities that help you instrument, evaluate, and iterate on your AI apps faster than ever—whether you're working solo or with an AI coding assistant by your side.

When we started building TruLens, our mission was clear: make it easier to evaluate, trace, and improve LLM-based applications so they can be trusted in production. Reaching 3,000 GitHub stars is a meaningful milestone, and we are genuinely grateful to everyone in the community who has contributed i

Agents are rapidly gaining traction across AI applications. With this growth comes a new set of challenges: how do we trace, observe, and evaluate these dynamic, distributed systems? Today, we’re excited to share thatTruLens now supportsOpenTelemetry(OTel), unlocking powerful, interoperable observab

In TruLens 1.1, we re-imagined the dashboard with a focus on making it easy to track large numbers of experiments, make comparisons and improve your apps for production. We also made several improvements performance and usability.

It has always been our goal to make it easy to build trustworthy LLM applications. Since we launched last May, the package has grown up before our eyes, morphing from a hacked-together addition to an existing project (trulens-explain) to a thriving, agnostic standard for tracking and evaluating LLM