Why Most Enterprise Software Isn't Ready for Agentic AILearnings From the Trenches Every enterprise software company is asking some version of the same question: How do we use AI agents to ship faster and unlock new customer value without putting our missMay 31 · 11 min readblog.codesweep.ai
Mixture of Open-Weight Models with Iterative Patch Generation Improves Performance on SWE-benchGoal Our goal in this study was to explore whether a mixture of open-weight models, combined through an iterative process, can outperform any single model on the SWE-bench Verified benchmark. Specifically, we wanted to evaluate if patches generated b...Dec 9, 2025 · 5 min readblog.codesweep.ai
Analysis of Reasoning Trajectories - Comparing Closed Weight Models vs Open Weight Models - Claude Sonnet 4 vs Kimi K2 InstructAbstract This study presents a comprehensive analysis of SWE-agent trajectories comparing Kimi K2 Instruct and Claude Sonnet 4 performance on software engineering tasks from the SWE-bench dataset. Through detailed examination of action category distr...Aug 5, 2025 · 6 min readblog.codesweep.ai
