AI Digs For Insights From Unstructured Data With Philippe Dame of Recollective

Subscribe at Thisnewway.com to get the step-by-step AI workflows. In episode 3 of This New Way, Aydin sits down with Philippe Dame, CEO of Recollective, to unpack how AI is transforming qualitative market-research, sales operations, internal knowledge-sharing, and day-to-day productivity.Phil walks through the product pivot that turned Recollective’s vast unstructured data into an “AI-first” insights engine, then demos a stack of no-code automations — from AI-powered lead scoring in Google Sheets to a self-serve Vertex AI search that makes every doc, deck, and Gong call instantly searchable.You’ll hear concrete play-by-plays, see cost breakdowns , and learn how to start small, scale fast, and turn AI into a force-multiplier across your org. Timestamps: 00:35 Phil’s background and early web-tool days 01:44 Pivot to Recollective and focus on market-research tech03:18 Why hybrid qual + quant research matters 03:58 Unstructured data → perfect AI playground 04:50 “Roadmap reset” after ChatGPT pressure 06:00 How the team up-skilled on LLMs & vector DBs 08:12 First AI features: summarization → theme extraction → comparisons 09:05 Making space for AI work without derailing commitments 10:10 Company-wide AI wins: sales data mining & lead scoring 13:08 Relay + Google Sheets workflow for automated enrichment 16:18 Running internal “AI office hours” to drive adoption 17:05 Staying current: newsletters, trials, and cost control 20:25 Seat-based vs. usage-based pricing—Phil’s take 23:04 Perplexity as Phil’s go-to research sidekick 24:50 Cutting the “collaboration tax” with self-serve AI answers 27:24 Live demo: Recollective’s Ask-AI tab & verbatim citations 31:28 Segment @mentions for instant comparative analysis 37:04 Emotion tagging and drilling into negative feedback 38:30 Building an internal Vertex-AI search in one afternoon 42:23 Agent Builder setup walkthrough 44:34 Relay use-cases: lead workflows, news scraping, stand-up bot 53:23 n8n migrations: 4,000 Gong calls plus on-the-fly analysis 56:30 OpenAI Playground & Notebook LM for ad-hoc knowledgebases 1:02:20 Google AI Studio multimodal experiments (free) 1:04:04 Start simple—one use-case, one stakeholder, iterate Tools & Resources Mentioned: AI / LLMs & Model-PlaygroundsChatGPT (OpenAI)OpenAI Platform & PlaygroundClaude (Anthropic)Perplexity Pro- Google Gemini (Workspace)Google AI Studio KnowledgeBase & Search Vertex AI Agent Builder / Search (Google Cloud) Notebook LM (Google) Internal Vertex-powered “ask-AI” portal Automation & WorkflowsRelay (no-code workflow tool with AI steps)Zapier (reference point)n8n (open-source automation + AI agents) Data, CRM & Sales EnablementGoogle SheetsSalesforceOutreachGong (call recordings) Collaboration & ProductivitySlack (AI channel, human-in-the-loop flows)Notion (company intranet)Google Drive & Google CalendarRSS feeds + custom “Article Extractor” scraper Core ProductRecollective (Phil’s qualitative-research platform with built-in AI features) Developer / Engineering UtilitiesCursor IDE (AI-assisted coding, briefly cited)