new file mode 100644
index 0000000..2109401
@@ -0,0 +1,26 @@
+---
+type: idea
+title: universal agentic searcher
+tags: [search, ai, agents, software]
+status: explored
+first_captured: 2025-10-16
+updated: 2026-04-10
+sources:
+ - sources/apple-notes/archived/universal-agentic-searcher.md
+---
+
+# universal agentic searcher
+
+a searcher that goes beyond what any single tool does for finding people, companies, events, communities, opportunities, and colleges. current approaches all have gaps: LLMs directly have outdated info, web search misses things, deep research can not avoid duplicates across runs, manual searches cannot cover all permutations, and blog lists are terrible.
+
+the proposed solution: a deep research pipeline (10+ steps, like a human researcher) with adaptive features — show preliminary results for feedback, time selector (1m, 5m, 20m), source prioritization, real-time research dashboard. output as a nice spreadsheet or any desired format. infrastructure: n8n + Apify + OpenAI.
+
+existing competitors: Datablist (sucky search), Clay (only GTM data), various scrapers (Import.io, Apify, Browse.ai, Thunderbit, Kadoa), Clado (YC, doing "deep research"), GPT-for-sheets. none combine the full pipeline. "still a bit clunky to implement (would like it though)."
+
+connects to the [[life-search|life search]] for personal data and to the [[quality-search|quality content search]] for the content-filtering variant.
+
+---
+
+## timeline
+
+- [2025-10-16] detailed competitive analysis, architecture design, gap analysis