Ask HN: How are you LLM-coding in an established code base?
Here’s how we’re working with LLMs at my startup. We have a monorepo with scheduled Python data workflows, two Next.js apps, and a small engineering team. We use GitHub for SCM and CI/CD, deploy to GCP and Vercel, and lean heavily on automation. Local development: Every engineer gets Cursor Pro (plus Bugbot), Gemini Pro, OpenAI Pro, and optionally Claude Pro. We don’t really care which model people use. In practice, LLMs are worth about 1.5 excellent junior/mid-level engineers per engineer, so p
阅读原始来源
来源记录
指标快照
points · 70 votes
2026-11-05 16:03:25.932Z
cumulative since publication
研究分析
用户报告的问题
来源描述：Here’s how we’re working with LLMs at my startup. We have a monorepo with scheduled Python data workflows, two Next.js apps, and a small engineering team. We use GitHub for SCM and CI/CD, deploy to GCP and Vercel, and lean heavily on automation. Local development: Every engineer gets Cursor Pro (plus Bugbot), Gemini Pro, OpenAI Pro, and optionally Claude Pro. We don’t really care which model people use. In practice, LLMs are worth about 1.5 excellent junior/mid-level engineers per engineer, so p
频次与替代方案
记录作者：adam_gyroscope。 来源票数: 70 votes (2026年11月5日 UTC; cumulative since publication)；前次同口径记录为 70（2026年11月5日 UTC），两个观测值不等于需求增长。 可供比较的近期同类记录：“MimiModel.encode: last frame differs from kyutai's encoder when the input length is not a multiple of 1920 samples” (GitHub Issues, 2026年11月5日)；“[问与答] iOS 从什么时候开始放弃长按拖动光标的？” (V2EX, 2026年11月5日)。主题相近仅用于导航，未核验其是否属于同一需求或独立证据。
需求判断
将报告的任务或障碍作为访谈线索。功能请求、抱怨或替代方案搜索本身并不能确定预算负责人或付费意愿。
商业证据缺口
待补证据：受影响的独立用户、重复发生场景、当前绕行方案、实际时间或金钱成本、采购权限，以及对具体报价的响应。
需求验证动作
按相同任务整理原始问题，去重作者和转载；核对今天如何处理、实际成本多少，再测试明确报价后判断付费需求。