MARS · The graph vs. the prompt

Four questions your AI can't answer alone

Wealth managers already research companies with Claude and Grok over web search and filings. That's excellent for one large-cap where the facts are concentrated. It breaks in the small- and micro-cap tail — where information is diffuse, names are ambiguous, and the pattern is hidden by design. Same company, same AI, four real questions. Left: a strong LLM + retrieval. Right: the same AI with the MARS graph as a tool.

The company: Conexeu Sciences, Inc. · NASDAQ: CNXU — a preclinical, zero-revenue biotech, shares +116% YoY, raising on a hot theme. A member asked a wealth manager to look at it. Here is that diligence, done both ways. Every figure on the right is a live query against the resolved graph.
01 · Entity resolution
"What else has this director — David Bogart — been involved in?"
The most dangerous failure: a confident, wrong answer.
LLM + Retrieval

"David Bogart is the Chief Executive Officer of Burford Capital, the litigation-finance firm."

CONFIDENTLY WRONG

That's Christopher Bogart. A different man. The model matched a name and fabricated a career. Nothing in the answer signals the error.

Same AI + MARS graph
Bogart, David Robert · CIK 2128417
resolved entity → edges: Conexeu Sciences only
Christopher Bogart (Burford) = separate node
RESOLVED · NO FALSE LINK

The graph resolved the real David R. Bogart by his SEC identifier and kept him distinct from the Burford CEO. It refuses the link the LLM invents.

Retrieval matches text; it cannot know two "Bogart"s are two people — or that two "Jeff Sharpe"s in the same filing set are one director and one namesake (the graph caught that too). For a regulated advisor, a confidently-wrong answer isn't useless — it's liability.

02 · Multi-hop over a diffuse pattern
"Do CNXU's 14 directors have a history of failed companies?"
The answer exists only across filings, never inside one.
LLM + Retrieval

14 directors × their prior companies × each outcome = 30–40+ retrieval hops. The model exhausts its context, resolves namesakes wrongly (see Q1), and misses any company not co-mentioned in a single document.

PARTIAL & UNRELIABLE

Returns a plausible narrative that can't be trusted or reproduced run-to-run.

Same AI + MARS graph
One traversal · directors → other cos → outcomes:
Harrison → Castle Biosciences, Novaestiq (acquired)
Pilcher → Suneva (acquired), Castle Biosciences
Costa → RX3 · Purcell → OnePointTwo Capital
→ 0 delistings · 0 bankruptcies · 0 SEC actions
COMPLETE · SOURCED

The pattern is diffuse by design — a promoter spreads across shells so no single filing shows the trail. Retrieval fetches documents; only a resolved graph fetches the pattern across them, in one hop, with a citation on every edge.

03 · The reverse screen
"Which OTHER companies share a board member with CNXU?"
The question that turns research into a pre-screen for your whole book.
LLM + Retrieval

Cannot be answered at all. There is no document that lists "companies sharing a director with CNXU." Retrieval can only return what you can name — it can't start from a pattern and find the matches.

NO DOCUMENT TO RETRIEVE
Same AI + MARS graph
Castle Biosciences (NASDAQ: CSTL)
  shares 2 directors — Harrison + Farberg
Suneva · Cytrellis · Servier · Sinclair …
  (Pilcher's network)
COMPUTED FROM STRUCTURE

A public company shares two board members with a promoted micro-cap — a board interlock no filing states.

This is the flip from "research this one" to "flag the three names in my 500-holding book I should look at." The answer is computed from the graph's structure, not retrieved from prose. No LLM + RAG can do it — at any number of turns.

04 · The claim of absence
"Is there any red flag in management's track record?"
Sometimes the finding is that there is nothing to find.
LLM + Retrieval

"I found no evidence of prior failures." But absence in the retrieved documents is not absence in the world. The model can report what it saw; it can never assert what isn't there.

CAN'T ASSERT ABSENCE
Same AI + MARS graph
Checked every prior board across all 14 directors:
zero delistings · zero bankruptcies · zero SEC actions
→ red flag is company-level (zero rev, +116%
  shares, promotion) — NOT a management failure-trail
POSITIVE COMPLETENESS CLAIM

"No prior track record" is itself the signal — first-time directors fronting a promoted shell. Only a resolved, bounded universe can turn silence into a defensible statement. The honest read — credible team, company-level red flag — is one the LLM alone could not have reached.

The pattern

Resolution

Knows two names are one entity — and one name is two people. RAG matches strings and confabulates.

Multi-hop

Fetches the pattern across filings in one query. RAG fetches documents and runs out of turns.

Reverse screen

Starts from a pattern and finds every match. RAG can only return what you can already name.

Absence

Asserts what isn't there, from a bounded universe. RAG can only report what it happened to retrieve.

The honest boundary: for a single large-cap — "summarize Apple's board" — the LLM with retrieval is genuinely fine, and the graph adds little. The graph is decisive exactly where information is diffuse, entities are ambiguous, and patterns are hidden — the small- and micro-cap tail. That the graph's edge and the market you're chasing are the same shape is not a coincidence. It's the reason.

And the moat follows: Claude and Grok improve with every model release — a rising tide that lifts your competitors' retrieval too. The graph improves as it accumulates resolved entities and verified edges — the one asset no model release hands anyone.