Blogs
Our technical challenges, ideas, how we're building them, and what we're learning along the way.
π A better Python cache for slow function callsβ
π§ Our simple vector database implementationβ
π Releasing our Open Source LLM Framework + Playground!β
π Large Language Models are Bad at Refactoring Codeβ
πΈοΈ A Simple Proxy for Azure and OpenAI raised our GPT4 TPM limit by 24xβ
π§ͺ Having GPT-4 Iterate on Unit Tests like a Humanβ
π Zero Downtime Deployment with < 50 lines of bashβ
πΊοΈ How an AI plans code changes across an entire GitHub Repositoryβ
π οΈ AI Junior Dev meets Super-Linterβ
π€ Why getting GPT-4 to modify files is hardβ
πΈ When you should handle tech debtβ
π Self Hosting an AI Junior Developerβ
π Improving LlamaIndexβs Code Chunker by Cleaning Tree-Sitter CSTsβ
π How our AI junior dev reads all of your documentationβ
π§Ή Sweep's Core Algorithmβ
βοΈ Generating 50k+ embeddings in 25 seconds using GTE π§ and MapReduce π»β
πͺ Chunking 2M+ files a day for Code Search using Syntax Treesβ
π€ Letting an AI Junior Dev run GitHub Actionsβ
π Code Search Infra for an AI junior developer - that doesn't store codeβ
π Understanding your codebase with ctagsβ
ποΈ Building a code search engine in one dayβ
π GPT 32k, π» Open-source, DeepLake ποΈ, GPT Functions Search π and more!β