What ClickHouse learned from a year of coding with AI agents
ClickHouse has been utilizing AI coding agents for a year, leading to significant insights about their effectiveness. Initially met with skepticism, the agents have proven useful for various coding tasks, especially after the release of Claude Opus 4.5. The company anticipates that 2026 will bring further productivity gains as the tools continue to mature.
- ▪ClickHouse has implemented AI coding agents to assist with software development tasks.
- ▪The effectiveness of these agents has improved significantly since the introduction of Claude Opus 4.5.
- ▪Agents have shown particular strength in handling repetitive tasks, resolving merge conflicts, and conducting code reviews.
Opening excerpt (first ~120 words) tap to expand
$(document).ready(function() { window.sponsorPortcoHits = window.sponsorPortcoHits || []; var sponsorSlug = 'clickhouse'; var postTitle = 'What ClickHouse learned from a year of coding with AI agents'; var url = 'https://thenewstack.io/clickhouse-ai-coding-agents/'; var categories = 'ai-agents,ai-engineering,software-development'; /// Prevent duplicate hits if (window.sponsorPortcoHits.includes(sponsorSlug)) { console.log('[sponsor_portco_hit] duplicate prevented for ' + sponsorSlug); return; } else { window.sponsorPortcoHits.push(sponsorSlug); console.log('[sponsor_portco_hit] unique hit detected for ' + sponsorSlug); var payload = { 'event' : 'sponsor_portco_hit', 'sponsor' : sponsorSlug, 'post_title' : postTitle, 'url' : url }; if (categories) { payload.categories = categories; }…
Excerpt limited to ~120 words for fair-use compliance. The full article is at The New Stack.
Discussion
0 commentsMore from The New Stack
