DraftlizeVOL. 1 · 2026 EDITION
Start free →
Use case · Investment memos

Investment memos that
show their work.

For analysts building investment cases. Source → fact → claim → decision chains where every claim cites its underlying observation, and decisions auto-flag stale when an assumption moves.

The standard analyst stack: where it leaks.

I

Sources buried in docs

You quote a 10-K figure in your memo. Three weeks later, partner asks "what was the source?" You search 47 PDFs. Maybe find it. Maybe not.

II

Claims drift from facts

You wrote "TAM $4B" in week 1 based on Q3 segment data. Q4 data revises TAM to $3.1B. Your memo still says $4B. Nobody noticed.

III

Conviction without provenance

You write "BUY PT $85." What chain of reasoning got there? In a standard memo it's prose; in three weeks you can't reconstruct which observations the price target rests on.

The Draftlize analyst loop.

I

Sources as observations

Drop 10-K, earnings call transcript, sector report into _refs/. Agent reads them, creates observation cards for each cited figure with →derived-from [[10-K-q3-2025.pdf]] wikilinks. Every number traces back.

II

Facts and claims as cards

c-04 fact: gross margin 38% Q3 2025 citing 3 observations. c-08 claim: GM expansion to 45% by FY27 with →supports c-04, c-12, c-15. Claim cards live alongside facts but are clearly different epistemic status.

III

Decisions with explicit chain

c-20 decision: BUY PT $85 with →derived-from a tree of claims, each supported by facts, each citing observations. The chain is queryable. Change a fact → claim flags stale → decision flags stale. Partner audit trail is automatic.

Real workflow: tracking a portfolio company over 6 months.

M1

Initial memo

~60 cards across observations / facts / claims / decisions. Original BUY PT card has a 4-level chain of provenance underneath it. Partner approves on Friday.

M3

Q4 earnings drops

You update 3 observation cards (revenue, GM, churn). Stale propagation flags 12 dependent cards across the chain — including the BUY PT decision. You see exactly which assumptions moved and re-derive the conclusion deliberately.

M6

Cross-portfolio conflict

Your memo says "competitive moat from network effects." Your colleague's memo on a competitor says "no real moat in the category." Cross-project conflict detection (when enabled) files a gap card in both. Partner sees both, decides which is right.

The substrate forces you to be explicit about the chain. Every claim has cited facts. Every decision has derivable provenance. Partner reviews go from "trust me" to "click to inspect."

Try it on one memo.

New accounts get $5 freePay only for what you useBalance never expires

Your free $5 is enough to seed one full memo (~40-80 cards) and feel one earnings update re-trigger the chain. Most analysts notice the audit-trail value by day 3.

Start free with $5