Pith turns the filings, earnings, and broker notes you read into a cited wiki per ticker — so your coverage knowledge is queryable and every line traces back to the source.
An equity or credit analyst reads more than almost anyone: 10-Ks and 10-Qs, transcripts, broker notes, sector news, regulatory filings — across a dozen or more names. The reading is constant, but the knowledge it produces scatters across PDFs, tabs, and a notes file that goes stale the day after earnings. And in this job the source isn't optional: every number in a thesis or a note has to be defensible back to a filing or a transcript. Pith makes the coverage knowledge layer first-class. You bookmark while you read; Pith tags each save to the right company or sector (auto), builds a cited wiki per name where every claim links to the source it came from, and assembles a briefing before earnings or a desk meeting (on demand). The reading you already do becomes a sourced, searchable memory of your coverage instead of evaporating after each print.
What changes for you
Scenario 1
Walk into earnings prep already briefed
The night before a name reports, generate a briefing from the quarter's saved sources — the prior transcript, the sell-side previews, the sector read-through, the guidance everyone is fixated on. Text plus audio for the commute. You go into the call knowing what changed and what to listen for, instead of re-reading three months of notes at 6am.
Scenario 2
Write a thesis where every claim is citable
When the note is due, your company wiki already organises the reading by concept — margins, leverage, competitive position, regulatory exposure — and every line links back to the filing, transcript, or broker note it came from. You draft from a sourced summary, and when the PM asks 'where's that from?', the citation is one click away, not a half-hour through your history.
Scenario 3
Onboard new coverage without months of catch-up
When a name moves onto your sheet, you start reading into it — and the wiki builds as you go. Two weeks in you have a cited page on the business, the bull and bear case, the comparable set, and the open questions, organised by concept rather than buried in a folder of links. Picking up coverage compresses from a quarter of catching up to a couple of weeks.
Founder's note
I'm a consultant, not an analyst, but I built Pith for the same problem: reading at volume and never being able to find — or source — what I'd read when it mattered. The thing that makes Pith fit research work is that the wiki is the artifact and every claim links to where it came from. I built citation-first because I wanted to trust my own summaries, and that's exactly what a defensible note needs.
FAQ
Is every claim in the wiki actually sourced?
Yes — citation-first by design. Each line in a company wiki links back to the bookmark it came from, so you can open the filing, transcript, or broker note and verify the number before it goes in a thesis. The wiki is the artifact; the source link is always attached.
How is this different from NotebookLM or a Notion workspace per name?
NotebookLM works in batch uploads — you re-upload when something changes, and coverage moves constantly. Notion makes you author and maintain the page yourself. Pith captures continuously from your reading flow and builds the cited wiki per name automatically, so each name stays current without a curation tax.
Where is my data stored?
Frankfurt, Germany — EU-only residency, per-workspace isolation at the database level, full export anytime as Markdown or JSON. We do not train models on your content.
How does Pith handle confidential or material non-public information?
Pith treats everything you save as private to your workspace — isolated, stored in Frankfurt, never used for training. It doesn't classify MNPI for you; apply your firm's information-barrier and compliance policy to what you choose to save, exactly as you would with any research tool.
Does it work for a solo analyst, or only a team?
Both. A personal workspace with one tag per name is the solo-analyst pattern; shared knowledge spaces let a desk or sector team bookmark into a common workspace and see the same cited coverage. You can query either from Claude or ChatGPT via MCP, with citations.
Related
Last reviewed: 6 June 2026