Monologue #3 | Coding Harness | London Session

​Breakthroughs in Coding Harnesses

​We’re bringing together some of the most interesting builders, researchers, designers, and AI-native engineers in London to explore a new frontier: coding harnesses.

​As AI coding agents become more capable, the real breakthrough may not come from the model alone, but from the systems around it — the harnesses, workflows, evaluations, interfaces, memory layers, and orchestration patterns that let humans and agents work together more effectively.

​This event is for people experimenting with new ways to code with AI: from agentic development environments and prompt/programming workflows, to autonomous debugging, repo understanding, evals, planning systems, and human-in-the-loop tooling.

​Expect a focused room of people actively building, sharing ideas, and testing new approaches. The goal is simple: create the conditions for new collaborations, sharper thinking, and maybe the next big leap in how software gets made.

​For: AI engineers, founders, researchers, product builders, toolmakers, and anyone pushing the edge of AI-assisted software creation.

​Format: Informal talks, demos, discussion, and high-signal conversations.

The Monologue was about exploring the Future of AI x HCI with OpenCues.com

Join our Whatsapp group: Join WhatsApp Group




Agenda

19:00-19:30 Arrival and mingle
19:30-20:30 Demo
20:30-21:30 Hands on Coding


Ben Arntzen

​Ben Arntzen is an open-source developer and systems engineer with a deep background in distributed storage, high-performance computing, Python, automation, and practical infrastructure. He brings a hands-on builder’s perspective to the conversation, especially around how complex systems behave in the real world and what it takes to make AI tools reliable outside the demo. Public sources describe him as a distributed storage enthusiast, systems engineer, open-source developer, and author of Raptors With Hats.

​Ibragim Badertdinov

​Ibragim Badertdinov is a Lead ML Engineer at Nebius working on software engineering agents, coding-agent evaluation, and large-scale benchmark infrastructure. He is one of the researchers behind SWE-rebench, a benchmark and dataset effort focused on realistic, continuously updated, decontaminated software-engineering tasks. His work explores how to evaluate coding agents more rigorously, beyond headline leaderboard scores.

​Eins Jae H. Yang

​Eins Jae H. Yang is an AI behavior engineer and designer working at the intersection of AI systems, human interaction, and product thinking. Through Monologue AI+HCI, Eins helps create spaces where builders, researchers, designers, and operators can examine emerging AI products in depth — not as passive demos, but as live pressure tests of ideas, interfaces, and future directions. Public listings describe Eins as an “AI Behavior Engineer & Designer for AI & Human,” and show him co-hosting Monologue AI+HCI events in London.

​Ruiqi

​Ruiqi will be speaking on AI benchmarking: how we measure model progress, where today’s benchmarks fall short, and what better evaluation practices could look like as AI systems become more capable and agentic. Expect a discussion that goes beyond scores and leaderboards into the deeper question of what benchmarks actually capture — and what they miss