{"id":29697,"date":"2024-07-30T11:27:13","date_gmt":"2024-07-30T09:27:13","guid":{"rendered":"https:\/\/www.graviton.at\/letterswaplibrary\/ui-for-data-enrichment-with-llms-search\/"},"modified":"2024-07-30T11:27:13","modified_gmt":"2024-07-30T09:27:13","slug":"ui-for-data-enrichment-with-llms-search","status":"publish","type":"post","link":"https:\/\/www.graviton.at\/letterswaplibrary\/ui-for-data-enrichment-with-llms-search\/","title":{"rendered":"UI For Data Enrichment With LLMs + Search"},"content":{"rendered":"<p><!-- SC_OFF --><\/p>\n<div class=\"md\">\n<p>I build a system to enrich datasets I found myself doing this a lot with LLMs connected to search. ChatGPT can&#8217;t do it yet as it doesn&#8217;t &#8216;loop&#8217;. The functionality is basic, but it works well. You can upload a CSV, provide instructions in natural language, preview results for top X rows, process task for full dataset, download results as CSV.<\/p>\n<p>Example tasks I have done:<\/p>\n<p>  Check if information seems to be valid based on top few search results and return a boolean Write a description of a company using LLM (+ optionally search results) Re-assign categories based on LLM  <\/p>\n<p>Is this of interest to anyone? Comment if so and I&#8217;ll put it online and send you a private link. Currently it uses my OpenAI API key so I would need to modify to BYOK or add billing, which I won&#8217;t bother with unless there&#8217;s interest.<\/p>\n<\/div>\n<p><!-- SC_ON -->   submitted by   <a href=\"https:\/\/www.reddit.com\/user\/oacoleshill\"> \/u\/oacoleshill <\/a> <br \/> <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1efo7by\/ui_for_data_enrichment_with_llms_search\/\">[link]<\/a><\/span>   <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1efo7by\/ui_for_data_enrichment_with_llms_search\/\">[comments]<\/a><\/span><\/p><div class='watch-action'><div class='watch-position align-right'><div class='action-like'><a class='lbg-style1 like-29697 jlk' href='javascript:void(0)' data-task='like' data-post_id='29697' data-nonce='614a020375' rel='nofollow'><img class='wti-pixel' src='https:\/\/www.graviton.at\/letterswaplibrary\/wp-content\/plugins\/wti-like-post\/images\/pixel.gif' title='Like' \/><span class='lc-29697 lc'>0<\/span><\/a><\/div><\/div> <div class='status-29697 status align-right'><\/div><\/div><div class='wti-clear'><\/div>","protected":false},"excerpt":{"rendered":"<p>I build a system to enrich datasets I found myself doing this a lot with LLMs connected&#8230;<\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[85],"tags":[],"class_list":["post-29697","post","type-post","status-publish","format-standard","hentry","category-datatards","wpcat-85-id"],"_links":{"self":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/29697","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/comments?post=29697"}],"version-history":[{"count":0,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/29697\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/media?parent=29697"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/categories?post=29697"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/tags?post=29697"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}