SWE-Agents, Benchmarks and Evaluation

I lead the development of autonomous SWE-Agents for intelligent code generation, issue resolution, and software testing at IBM Research. I returned to blogging after a ten-year break. Through this blog, I share my observations from that work.