Labs
Jev made us take another look at tool search
How we used Jev to select tools for Conversion's agent, with benchmarks against lexical, BM25, and native tool search.
Sep 28, 2026 · 5 min read
September 30, 2026 6 articles
Labs
Research, experiments and benchmarks from the team building the platform.
Labs
One-to-One: Evaluating AI Personalization in Marketing Systems
One-to-One: a benchmark for AI agents building personalized marketing content.
Vikram Sivashankar September 28, 2026 6 min read
Labs (5)
Labs
Using semantic benchmarks to build a self-improving text-to-query agent
Our text-to-query agent went from 45 seconds on frontier models to 2 seconds on GLM 5.3 Flash, with the same accuracy at 1/20 the cost. This post describes the architecture, evaluation loop, and lessons learned.
Sep 2, 2026 · 10 min read
Labs
Building a data flywheel for our agent harness
An observability pipeline that reads production traces at corpus scale, names recurring failure modes, and proves the fix held.
Aug 22, 2026 · 7 min read
Labs
What we took out of the context window
A median request carried 150,000 tokens. Most of our context work since comes down to one move, applied repeatedly.
Aug 14, 2026 · 7 min read
Labs
From an agent harness to a research platform
How and why we made our harness model-agnostic, and what per-turn portability actually cost us.
Aug 6, 2026 · 7 min read