# BrainGraph

> AI digital brain platform for coaches and consultants. Turn client calls, documents, and frameworks into a personal knowledge graph that AI agents can actually reason with — built 1-on-1 with you, not as another self-serve PKM course.

BrainGraph is built and run by Dimitris Goudis. The product turns transcripts, frameworks, and client notes into a connected entity graph stored in FalkorDB and queried via the Model Context Protocol — so any AI agent (Claude, ChatGPT, custom) can prep meetings, surface client patterns, draft content, and answer questions in your voice.

## Core pages

- [Home](https://itsbraingraph.ai/): Overview of the AI digital brain — what it is, who it's for, how it works.
- [Pricing](https://itsbraingraph.ai/pricing): Free (40 docs total, no credit card), Pro ($49/mo, 60 new docs/month + all integrations including Zapier/Make/n8n/Claude Cowork/OpenClaw/API), Max (100 new docs/month; $199 today — 4-hour 1-on-1 build session with Dimitris plus 3 months of Pro included; then $49/mo, cancel anytime).
- [1-on-1 Consulting](https://itsbraingraph.ai/consulting): Build your AI digital brain with Dimitris in a single 4-hour 1-on-1 session, screen-shared. Discovery call is free. Bundled into the Max plan.
- [About Dimitris Goudis](https://itsbraingraph.ai/about): Founder bio — why BrainGraph exists, what Dimitris does today, and what he writes about.
- [Free Masterclass](https://itsbraingraph.ai/masterclass): Two free live sessions per week — Wednesday masterclass on building your digital brain, Friday office hours on AI agents and integrations.
- [API Documentation](https://itsbraingraph.ai/documentation): Public BrainGraph API reference. Endpoints for uploading documents (POST /api/documents/upload, POST /api/documents), ingesting public web pages by URL (POST /api/documents/url), ingesting YouTube videos (POST /api/documents/youtube), transcribing audio recordings and voice memos (POST /api/transcripts/upload-audio, up to ~150 MB), polling processing status (GET /api/documents/{id}/progress), listing the brains in a multi-brain account (GET /api/brains), listing and name-searching the documents you have uploaded with their upload dates (GET /api/brain/documents), natural-language brain search (POST /api/brain/search, with optional brainIds to narrow), and entity resolution (GET /api/brain/entity/resolve). Bearer API key with documents:write or brain:read scopes. GET /api/brain/capabilities returns a machine-readable index for AI agents.
- [Claude MCP Connector](https://itsbraingraph.ai/docs/claude-connector): How to install and use the Brain Graph Model Context Protocol (MCP) connector in Claude Desktop and claude.ai. Includes the install URL (https://mcp.itsbraingraph.ai/mcp), the ten built-in tools (search_brain, resolve_entity, get_document_images for pulling figures/diagrams from uploaded documents, upload_text, upload_to_brain for in-chat file uploads, upload_document, upload_url for web articles, upload_youtube_url for YouTube videos, upload_audio_url for voice memos and audio recordings, list_brains for multi-brain accounts), example prompts, security model, and beta status. Free on every Brain Graph plan.
- [Build Your Own Digital Brain System](https://itsbraingraph.ai/build-your-own-digital-brain-system): Consulting sprint (one to two weeks) that designs and builds a company's own domain-specific AI digital brain system — custom architecture from the whole company, AI agents built on top, real code, and a handover that teaches the client to run and improve it. For early-adopter consultants and experts who need depth a general-purpose tool can't reach. Differs from BrainGraph itself, whose analysis is general-purpose (transcripts, article URLs, YouTube, documents).
- [Glossary](https://itsbraingraph.ai/glossary): Plain-English glossary of the terms behind AI-powered knowledge — digital brain (BrainGraph's term for a knowledge graph an AI can reason with), second brain (the Personal Knowledge Management term for a notes archive), knowledge graph, graph database, node, edge, entity, fact, knowledge base, AI agent, agent orchestration, skill file, soul file, memory file, MCP, RAG, PKM, and ingestion. Every definition steers from 'second brain' toward 'digital brain.'
- [AI for Consultants](https://itsbraingraph.ai/for-consultants): Pillar page on how independent business consultants use an AI digital brain — built from their own client work — across four jobs: cold outreach, spotting patterns across clients, content from real conversations, and client meeting prep. Bridges the familiar 'second brain' (a notes archive) to a 'digital brain' (a knowledge graph an AI can reason with). Two entry points: start free with BrainGraph, or build a custom system in a sprint.
- [Cold Outreach With Your AI Digital Brain](https://itsbraingraph.ai/use-cases/cold-outreach): How to use an AI digital brain for cold outreach — score a profile against your ICP, investigate the person to understand what would be valuable to them, and draft a specific, human first message. The brain drafts; the consultant decides and sends.
- [Spot the Patterns Across All Your Client Work](https://itsbraingraph.ai/use-cases/client-patterns): How an AI digital brain surfaces common problems and patterns across all of a consultant's client work at once — ingest every conversation into one knowledge graph, ask across all of it, and turn recurring patterns into offers, content, or a sharper methodology.
- [Turn Client Conversations Into Content](https://itsbraingraph.ai/use-cases/content-from-conversations): How an AI digital brain turns real client conversations into distinctive content — the brain is built from the consultant's actual calls and notes, so a draft starts from a real example, objection, or win instead of the generic internet average.
- [Client Meeting Prep in Seconds With Your Digital Brain](https://itsbraingraph.ai/use-cases/meeting-prep): How an AI digital brain preps a consultant for client calls in seconds — every call, email, and note about a client is ingested and connected, so the brain can assemble recent history, open threads, and priorities into a pre-call brief on demand.
- [BrainGraph vs Delphi](https://itsbraingraph.ai/braingraph-vs-delphi): BrainGraph vs Delphi. Delphi builds a public-facing AI clone (a "digital mind") that answers as you to your audience across text, voice, and video. BrainGraph builds a private digital brain: a knowledge graph from your transcripts, notes, and documents that you and your AI agents (via MCP) query for meeting prep, outreach, content, and pattern-spotting. Delphi scales your presence outward to other people. BrainGraph scales your thinking inward for your own work. Both are trained only on your own material, and people with an audience often run both.
- [BrainGraph vs Coachvox](https://itsbraingraph.ai/braingraph-vs-coachvox): BrainGraph vs Coachvox. Coachvox is an audience-facing AI coach chatbot embedded on your website that engages visitors 24/7 and captures qualified leads (top of funnel). BrainGraph is a private digital brain: a knowledge graph from your transcripts, notes, and documents that you and your AI agents (via MCP) use for meeting prep, follow-ups, content, and pattern-spotting (the work after the lead). Coachvox faces your prospects. BrainGraph faces you. They sit at opposite ends of the same practice and are often used together.
- [BrainGraph vs Notion AI](https://itsbraingraph.ai/braingraph-vs-notion-ai): BrainGraph vs Notion AI: what each does and when to use it. Notion stores notes, documents, and long-context files in one place that humans edit and maintain, with an AI feature that answers questions by searching those pages on relevant terms. BrainGraph processes and analyses your content (about 20 minutes per document) into a knowledge graph on a graph database, extracting topics, problems, solutions, timelines, and statuses, storing the relationships between them, and weighting each piece so frequently-mentioned topics don't dominate (AI weighs the whole base evenly). The result is queried efficiently by AI and used as a knowledge source inside automations (e.g. an agent drafting a client email investigates the client in the brain, finds the core question and related problems, and writes an impactful reply). Both serve solopreneurs and teams; both offer a connector usable from tools like Claude Code. They pair well: edit in Notion, let BrainGraph process it into the AI brain. Teams wanting their own knowledge base may need a custom analysis workflow designed for their business, built in a sprint.
- [BrainGraph vs Obsidian](https://itsbraingraph.ai/braingraph-vs-obsidian): BrainGraph vs Obsidian. Obsidian is a local-first, plain-text Markdown notes app where you own your files and build a graph by hand with links; its graph view maps the links you typed, and it has no built-in AI (added via plugins). BrainGraph reads your real work (calls, transcripts, notes, documents) and automatically builds a knowledge graph of the people, problems, decisions, and dates and their relationships, which you and your AI agents query via MCP. Obsidian = a graph you build and own by hand, for you to explore; BrainGraph = a graph built automatically, for AI to reason over. They pair well: write and own notes in Obsidian, analyse everything into BrainGraph for AI.
- [BrainGraph vs NotebookLM](https://itsbraingraph.ai/braingraph-vs-notebooklm): BrainGraph vs NotebookLM. Google NotebookLM is a source-grounded research tool: you upload a fixed set of sources into a notebook and it answers strictly from them with citations, plus summaries, study guides, quizzes, and audio overviews. Best for understanding a defined body of material. BrainGraph is a persistent, evolving knowledge graph of your whole practice (calls, notes, documents) that your AI agents query via MCP to prep, draft, and decide. NotebookLM is scoped per notebook and built for you to chat with; BrainGraph is always-on, grows across everything, and is built for agents. They pair: research a fresh source set in NotebookLM, keep durable insights in BrainGraph.
- [Zapier Integration](https://itsbraingraph.ai/zapier): Hub page for BrainGraph's Zapier integration. Lists ready-made Zap templates that auto-ingest content into a user's AI digital brain: Google Drive files (with an export variant for native Google Docs), YouTube videos by search or by channel, and Krisp and tl;dv meeting transcripts. Brain Graph actions used are Upload Document (File), Upload YouTube Video, and Send Text Content.
- [Save Google Drive Files to Your Digital Brain (Zapier)](https://itsbraingraph.ai/zapier/google-drive-files-to-brain-graph): Zap template that watches a Google Drive folder (trigger: New File in Folder) and uploads each new file to BrainGraph via the Upload Document (File) action, with an optional Filter by Zapier step to limit to supported document types (PDF, DOCX, TXT, MD). For portable files already in a readable format.
- [Upload Google Docs to Your Digital Brain via Zapier](https://itsbraingraph.ai/zapier/google-docs-export-to-brain-graph): Zap template for native Google Workspace files. Trigger: Google Drive New File in Folder; an optional Filter by Zapier step; a Google Drive Export File step that converts Google Docs/Sheets/Slides to PDF or Word; then Brain Graph Upload Document (File). Use this variant when the folder holds native Google documents rather than already-portable files.
- [Save YouTube Videos by Search to Your Digital Brain](https://itsbraingraph.ai/zapier/youtube-videos-by-search-to-brain-graph): Zap template that monitors YouTube for new videos matching a search query (trigger: New Video by Search) and ingests each into BrainGraph via the Upload YouTube Video action — transcript plus metadata. Recommends narrow searches (creator names, exact phrases, niche topics). Each video is processed once.
- [Save a YouTube Channel's Videos to Your Digital Brain](https://itsbraingraph.ai/zapier/youtube-channel-videos-to-brain-graph): Zap template that captures every new upload from a chosen YouTube channel (trigger: New Video in Channel) and ingests it into BrainGraph via the Upload YouTube Video action — transcript plus metadata. Optional Filter by Zapier step to limit by title. Each video is processed once.
- [Save Krisp Meeting Transcripts to Your Digital Brain](https://itsbraingraph.ai/zapier/krisp-transcripts-to-brain-graph): Zap template that captures new Krisp meeting transcripts (trigger: Transcript Created) and sends them to BrainGraph via the Send Text Content action, which runs entity extraction on the plain text. Makes meeting decisions, commitments, and objections queryable.
- [Save tl;dv Meeting Transcripts to Your Digital Brain](https://itsbraingraph.ai/zapier/tldv-transcripts-to-brain-graph): Zap template that captures new tl;dv meeting transcripts (trigger: Transcript Added) and sends them to BrainGraph via the Send Text Content action for entity extraction. Stores speaker-by-speaker context so quotes, objections, and decisions from past calls are retrievable.
- [Features](https://itsbraingraph.ai/features): Full feature catalog for BrainGraph (itsbraingraph.ai). Every capability the digital brain supports today, grouped into Capture (voice/audio notes, file upload, paste text, URLs, YouTube, in-chat file upload, image OCR), Organise (multi-brain / sub-brains, doc-ready email notification, document image preservation), and Connect (search your brain, MCP connector for Claude, Zapier integration, public API). Each feature links to its dedicated page or the closest documentation.
- [AI Voice Notes](https://itsbraingraph.ai/voice-notes): AI voice notes feature in BrainGraph (itsbraingraph.ai). Record audio in the browser or upload a voice memo, get speaker-diarized transcripts in 100+ languages (powered by AssemblyAI), and have the recording folded into the user's digital brain alongside documents, URLs, and YouTube transcripts so questions like 'what did X say about Y' return cited answers. Up to 150 MB per file. Supported formats: mp3, m4a, wav, ogg, oga, opus, webm, aac, flac, mp4. Triggerable via MCP (upload_audio_url tool in Claude) and Zapier. Differentiator vs Otter/Fireflies: those are meeting bots that produce transcripts; BrainGraph is the layer underneath where voice notes, meeting transcripts, documents, and URLs all connect.
- [Multiple AI Brains](https://itsbraingraph.ai/multi-brain): Multi-brain (sub-brain) feature in BrainGraph (itsbraingraph.ai). One account can hold many separate AI knowledge graphs (e.g. Personal, Marketing, Learning, one per client). Each brain has its own entities and documents; they do not bleed across. Captures route via a brainId parameter on every upload path (web, API, MCP connector, Zapier). Search spans all brains by default and only narrows when the user names a brain ('search my Marketing brain'). One connection in Claude and Zapier covers every brain. No reconnecting. Plan limits count across all brains together. Designed for consultants juggling clients, coaches keeping personal and client work separate, and builders with parallel projects. Different from having multiple accounts: one login, one bill, one connection, segmented memory.
- [AI Briefings](https://itsbraingraph.ai/ai-briefings): AI Briefings v2 feature in BrainGraph (itsbraingraph.ai). A briefing is a Notion-style editable document the user's digital brain writes from their knowledge, not a one-shot summary. Pipeline: parseRetrievalBrief (Kimi via OpenRouter parses the user's description into a typed brief with anchors, intent axes, exclusions, temporal window, voice) → proposeSections (Kimi designs 3-6 sections) → retrieveCandidates (per-brain searchGraphV2 fan-out, aggregate ~60-100 candidates, dedup preserving brain attribution) → routeAndFilter (single Kimi call assigns each candidate to a section with confidence + reason; below-threshold dropped server-side) → synthesizeSection (parallel per-section, writes graph_card blocks grounded in cited fact UUIDs). The Notion-style editor (BlockNote) lets users edit, reorder, add their own blocks; citations are inline pills. A right-rail pending tray surfaces auto-update proposals when new documents land in selected brains, with accept/reject controls and confidence color-coding (green ≥0.85, amber ≥0.7, red below). Rejections feed briefing_exclusions so the same source never re-proposes. Accuracy guarantees by construction: section_id enum-constrained, factUuids subset-verified, exclusions stripped before routing. Multi-brain by design: one briefing can fan out across many brains with attribution preserved.
- [Blog](https://itsbraingraph.ai/blog): Articles on AI agents, digital brains, knowledge graphs, and packaging consulting expertise for AI.

## Blog posts

- [Which LLM Builds the Best Knowledge Graph? Opus 4.8 vs Fable 5](https://itsbraingraph.ai/blog/llm-knowledge-graph-benchmark-opus-vs-fable-5): We ran Claude Opus 4.8 and Fable 5 over the same 134 videos to build identical knowledge graphs, then verified every number against the transcripts. Here's which model reasons better — and what it means for your AI digital brain.
- [Will a Digital Brain Actually Make You More Money?](https://itsbraingraph.ai/blog/will-a-digital-brain-make-you-more-money): Not directly. But here's exactly how a digital brain adds money to a consulting practice, the math to estimate your own number, and where it does nothing.
- [Digital Brain for Business: Personal Productivity or the Whole Company?](https://itsbraingraph.ai/blog/digital-brain-personal-productivity-vs-business): Is a digital brain for founder productivity or the whole business? Both — here's how your personal brain becomes the business's brain, and how to own it.
- [How Boy Hijnen Turned a 127-Metric Consulting Framework Into a Running AI Agent](https://itsbraingraph.ai/blog/consulting-framework-to-ai-agent-client-yield): How Boy Hijnen turned a 127-metric consulting framework into a running AI agent using Neo4j + Claude — in 3 months, without being technical.
- [AI Agent Guardrails: How Memory and Soul Files Stop the Chaos](https://itsbraingraph.ai/blog/ai-agent-guardrails-memory-soul-files): AI agents pushed code to production without permission. Here's how memory files, soul files, and guardrails turned chaos into control — with real examples.
- [AI Skill Files: How to Package Your Expertise for AI Agents](https://itsbraingraph.ai/blog/ai-skill-files-package-expertise-for-ai-agents): Learn what AI skill files are, how to create them in 5 steps, and why consultants are packaging their expertise as portable SKILL.md files for AI agents.

## Legal

- [Privacy Policy](https://itsbraingraph.ai/privacy-policy): How BrainGraph collects, uses, and protects personal data. User content is private and never used to train AI models.
- [Terms of Service](https://itsbraingraph.ai/terms): Terms of service for BrainGraph — service description, content ownership, acceptable use, billing, and cancellation.
- [Cookie Policy](https://itsbraingraph.ai/cookie-policy): How BrainGraph uses essential, analytics, and embed cookies, and how users manage cookie preferences.

## About

- Founder: Dimitris Goudis — builds AI digital brains for coaches and consultants. Profile: https://www.linkedin.com/in/dimitris-goudis/
- Contact: dimitris@3nuggets.io
- Sitemap: https://itsbraingraph.ai/sitemap.xml