The reading memory for consultants

Everything you've read — ready before every client call.

Pith builds a cited wiki from what you read — organized by client, queryable by your AI agents. When it counts, it delivers the work: briefings, dossiers and audio where every claim links a source you vetted.

Start free See the live demo

No credit card, nothing to cancel. Deliverables at fixed prices: briefing €2 · dossier €5 · audio €3.

Data in Frankfurt
GDPR-compliant
No training on your content

Articles you save, distilled into concepts

Find Pith your way

Switching tools, or after an AI research assistant that remembers what you read?

However you got here, the door is the same: leave a tool that makes you author everything, or give your AI a living memory instead of project boxes — built while you read, every claim traceable to a source you vetted.

AI research assistants Notion alternative Obsidian alternative NotebookLM alternative Read-it-later alternatives

Who's behind it

„I build Pith as a tech consultant for my own work — and I'm customer number one. What you see here, I use every day myself."

Stefan

Founder of Pith

A Pith wiki concept page on Retrieval-Augmented Generation with cited sources — Real Pith UI — a wiki page Pith wrote from saved articles, every claim cited to its source. No mockups.

What Pith is for

You read more than you remember. And generate more than you can prove.

Both get expensive in the same moment — in front of the client, when someone asks “says who?” Pith builds a living client memory from what you read: cited, searchable, ready when it counts.

When sources disagree, Pith tells you

Two articles, two claims: 'DORA deadline Q1' vs. 'pushed to Q3'. Pith catches the contradiction and shows you both sides — with sources. No chatbot does that.

A briefing before the meeting

When it counts, the memory produces a cited meeting briefing — €2, price on the button. Failed jobs cost nothing.

A wiki that writes itself

Your reading becomes topic pages with clickable sources. You maintain nothing — it's part of the free workshop.

Your AI reads along

Ask Claude or ChatGPT what you already read — over MCP, free, with your agent's own tokens. Answers come with the source.

Save in one click

Browser extension, phone, or RSS — articles in, summarised automatically. Free, forever — 200 sources a month included.

How it works

From bookmark to briefing in four steps.

Save

Browser extension, mobile share-sheet, or RSS. The article is fetched, cleaned, and queued for processing.

Highlight

Mark passages as you read. Your highlights land in the wiki — searchable, citeable, and they resurface when they're relevant again.

Tag a client

Optional. Tag bookmarks with the engagement they belong to. Briefings, search, and the wiki then filter to that client's context.

Buy the deliverable

When it counts, the memory produces the work: the meeting briefing before the meeting (€2), the client dossier before the project (€5), the audio radar for the ride (€3). Price on the button; failed jobs cost nothing.

Real output

From your bookmarks to a wiki page.

Three real wiki pages Pith generated from articles you'd save during a normal consulting week. Every claim links back to a source. Click a badge to jump. Switch tabs to compare domains.

What you saved

Articles bookmarked across the week.

1
DORA Article 30: Key Contractual Requirements
securiti.ai
2
DORA – Managing of ICT third-party risk
fma.gv.at
3
Digital Operational Resilience Act, Article 30
digital-operational-resilience-act.com
4
EU DORA Requirements for ICT Service Providers
cm-alliance.com
5
DORA Third-Party Risk: Register of Information & CTPPs 2026
regulation-dora.eu
6
DORA Article 30: ICT Contract Requirements for Financial Entities
regulativ.ai

What Pith built

Auto-generated · 18 May 2026

Digital Operational Resilience Act Article 30

How EU financial entities must contract with ICT third-party providers under DORA.

The Digital Operational Resilience Act (DORA) Article 30 mandates specific contractual provisions that financial entities must embed in agreements with ICT third-party service providers to ensure operational resilience and risk management.1 3 4 5

Key Features

Article 30 requires contracts for critical or important ICT functions to include over 15 mandatory clauses covering governance, risk mitigation, and operational resilience.2 4 5 Contracts must clearly define service descriptions, sub-contracting permissions, service locations, data-protection obligations, and incident-reporting duties.1 4

Risk Management and Compliance

The contracts must define each party's responsibilities, set service levels, and establish monitoring mechanisms for ICT risks.1 3 4 Financial entities retain audit rights — including direct access to provider premises — and termination rights with exit strategies that preserve business continuity.2 5

Why It Matters

Negotiating and enforcing these clauses is non-trivial with large ICT providers, but embedding them is the cornerstone of DORA compliance. The Register of Information cycle running through Q1 2026 makes contract gaps visible to supervisory authorities for the first time.5 6

Sources6 sources

1
DORA Article 30: Key Contractual Requirements for Operational Resilience — Securiti
https://securiti.ai/dora-article-30/
2
DORA – Managing of ICT third-party risk — FMA Österreich
https://www.fma.gv.at/en/cross-sectoral-topics/dora/dora-managing-of-ict-third-party-risk/
3
Digital Operational Resilience Act (DORA), Article 30
https://www.digital-operational-resilience-act.com/Article_30.html
4
EU DORA Requirements for ICT Service Providers
https://www.cm-alliance.com/cybersecurity-blog/eu-dora-requirements-for-ict-service-providers-all-you-need-to-know
5
DORA Third-Party Risk: Register of Information & CTPPs 2026
https://www.regulation-dora.eu/third-party-risk
6
DORA Article 30: ICT Contract Requirements for Financial Entities — Regulativ.ai
https://www.regulativ.ai/blog-articles/dora-article-30-a-strategic-guide-for-ict-providers-and-financial-entities

Explore the live demo →

Same pages are exposed via MCP — your AI assistant cites these sources too.

Your agent, via MCP

The cited memory your AI answers from.

Your agent thinks and writes — Pith remembers and proves.

Connect Claude, ChatGPT, or your own agent over MCP and it reads your memory free, with its own tokens: it searches your bookmarks, opens the cited wiki pages, and answers with the sources you vetted. No other vendor serves its memory to someone else's agent — which is exactly why Pith does.

Features

See what your reading turns into.

Five things that set Pith apart from a notes app, a bookmark manager, or a chat-with-your-PDFs tool.

Auto-generated wiki page on Retrieval-Augmented Generation, with cited paragraphs and a related-pages sidebar.

Your reading writes itself into a wiki.

Concepts emerge from what you save — people, frameworks, regulations, technologies. Each page is built by an LLM from your bookmarks, with citations linking back to the source. No authoring required.

Learn more

Topic-map view with five color-coded clusters: AI & transformers, Rust, React, Database, DevOps.

See how your knowledge connects.

A force-directed graph of every bookmark you've saved, clustered by what they're actually about. Zoom out for the map; zoom in for individual cards. Surfaces connections you didn't notice.

Learn more

Auto-tag review queue with pending suggestions grouped under FinTech AG and Anthropic, each with a confidence percentage and confirm/dismiss buttons.

Bookmarks tag themselves against your accounts.

Pith scores every save against your client roster using entity matches, domain affinity, and embedding similarity. High confidence applies the tag automatically; mid confidence waits in a queue you clear in seconds.

Learn more

Morning Briefing surface with a 10-minute audio digest call-to-action.

A cited briefing before every meeting.

One click, €2, price on the button: a per-client briefing from the sources you saved and vetted — text plus narrated audio for the way in. Every claim links its source. Failed jobs cost nothing.

Learn more

Acme Consulting client detail page with auto-tag settings panel and a list of recent client-tagged bookmarks.

A knowledge folder per client, kept fresh.

Tag bookmarks with clients to scope your wiki, search, and feeds to one engagement at a time. Each client gets an activity stream, briefings, and an auto-tagging settings panel.

Learn more

View all features →

YOUR JUDGMENT

Your assessment, attached to the source.

Add your own notes to any source. Saved becomes saved → assessed: your judgment beside the source, not just a pile of links.

why it matterswho it's forwhat supersedes it

From the Pith blog

Notes on knowledge work

Field notes from building a reading memory. Updated weekly.

All posts →

Pith

Knowledge ManagementSep 21

When your knowledge base is wrong about itself

A machine-maintained knowledge base has a failure mode a handwritten one does not: it can be wrong about itself and look entirely fine. What we found auditing our own wiki — and what follows from it.

PPith Lab Team

Pith

Knowledge ManagementSep 20

Where did I read that? Every mention, in one place

DORA turns up in four things you read across three months. You remember two. Pith recognises the named things in everything you save — regulations, institutions, companies, people — and gathers every mention in one place.

PPith Lab Team

Pith

Knowledge ManagementJun 7

Your reading, as one knowledge graph

You don't read in silos, so why store reading in them? Pith distils everything you save into one knowledge graph — and the adjacencies it surfaces are where the insight lives.

PPith Lab Team

FAQ

Common questions.

Who is Pith for?

Consultants, analysts, and small consulting firms. Anyone whose work depends on reading widely and re-finding what they read.

Why not just ChatGPT?

Use ChatGPT — Pith assumes you do. What your chat doesn't have is your reading: it doesn't know what you saved, sometimes invents sources, and forgets everything when the chat ends. Pith is the memory underneath: your agent queries it via MCP and answers with citations to sources you actually vetted. Your agent thinks and writes — Pith remembers and proves.

How is this different from Notion or Confluence?

Notion and Confluence are for writing things yourself. Pith writes itself: it remembers what you read, summarises it, and builds a wiki with sources — without you maintaining any pages.

Is my data isolated from other workspaces?

Yes — per-workspace database. Cross-workspace reads are impossible at the SQL layer, not just access-controlled.

Do you train models on my content?

No. Your content is used to generate your briefings, your wiki, your search results — and nothing else.

Can I export my data?

Yes, anytime. One-click export as a Markdown bundle + JSON. Your data is yours.

Does Pith work with Claude, Cursor or claude.ai?

Yes. Pith exposes your cited knowledge over MCP, so your AI assistant can query it directly — with source citations, no copy-paste into the prompt.

Save for free. Pay per deliverable.

No credit card, no subscription — nothing to cancel. Your knowledge stays yours.

A €2 briefing before a client meeting is a project cost, not a software budget — attributed to the engagement, re-billable via the monthly statement. Failed jobs cost nothing; your balance never expires.

Start free