Top all-time · #14477
Daft
Company blog · daft.ai/blogdataai
Daft is the company blog at daft.ai, writing about data and ai. Since 2025, 5 different Hacker News readers have linked it 17 times, most often “Knowledge curation (not search) is the AI big data problem”. No feed was found, so read it on the site itself.
Too few sampled comments (1) to say how often HN recommends it.?
Why HN links it
Linked in an HN thread on giving models continuous context. The only comment discusses agentic search and how models gather context, citing Anthropic's Claude Agent SDK post rather than this blog.
Summarised by Claude from a sample of 1 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Daft is a high-performance data engine for AI workloads. Process any modality at any scale with Python DataFrames and SQL — from your laptop to petabyte-scale distributed clusters.
Most-linked posts?
- Knowledge curation (not search) is the AI big data problem2 people · 3 mentions · first linked 2025-12-24
- Cutting LLM Batch Inference Time in Half: Dynamic Prefix Bucketing at Scale2 people · 2 mentions · first linked 2025-11-04
- Embedding Text Documents with Qwen32 people · 2 mentions · first linked 2025-08-13
What Hacker News says?
It's fairly surprising to me how naive/early we are still in the techniques that we use here. Anthropic's post on the Claude Agent SDK (formerly Claude Code SDK) talks about how the agent "gathers context", and is fairly accurate as to how people do it today. 1. Agentic Search (give the agent tools and let it run its own search trajectory): specifically, the industry seems to have made really strong advances towards…
Comment linking /blog/knowledge-curation-not-search-is-the-big-data-problem-for-ai
on “Ask HN: What is best way to provide continuous context to models?”Submitted Embedding Text Documents with Qwen3 as a story · 18 points
HN discussionSubmitted We Hit 100% GPU Utilization–and Then Made It 3× Faster by Not Using It as a story · 17 points
HN discussionSubmitted Cutting LLM Batch Inference Time in Half: Dynamic Prefix Bucketing at Scale as a story · 5 points
HN discussionSubmitted Agentic systems are just query engines for unstructured data as a story · 3 points
HN discussionSubmitted Knowledge curation (not search) is the AI big data problem as a story · 2 points
HN discussionSubmitted How We Use AI Coding Agents as a story · 2 points
HN discussionSubmitted Cutting LLM Batch Inference Time by Half with Dynamic Prefix Bucketing as a story · 2 points
HN discussionSubmitted Show HN: Flotilla, the Distributed Engine for Multimodal Pipelines as a story · 2 points
HN discussionSubmitted Scaling As-Of Joins as a story · 1 points
HN discussionSubmitted VLAs are dead, long live World Action Models as a story · 1 points
HN discussionSubmitted Search is not the AI big data problem as a story · 1 points
HN discussionSubmitted How We Use AI Coding Agents as a story · 1 points
HN discussionSubmitted Processing 99% of U.S. Caselaw for Under $1 as a story · 1 points
HN discussionSubmitted Benchmark: Spark vs. Ray Data vs. Daft on Multimodal Workloads as a story · 1 points
HN discussionSubmitted Profiling multimodal workloads: lessons from Daft as a story · 1 points
HN discussionSubmitted Processing 24T tokens for LLM training with 0 crashes (what made it possible) as a story · 1 points
HN discussion
