Daloopa

The removal test leaves an archive that decays immediately. Models read filings, footnotes, investor presentations and transcripts and turn them into structured, source linked datapoints, and the defining use case is keeping a model current within hours of a company reporting, which only works if extraction runs continuously.

Strip the models and the accumulated dataset remains as a historical record while the product itself stops existing, because earnings season is the moment the tool earns its place. The company was founded by a buy side analyst who describes the problem as manually sourcing and entering data taking time away from actually analysing investments, so displacing that manual work is the entire premise rather than a feature on top of one.

Autonomy exposure is structurally low because the product supplies data rather than making decisions, and the analyst who receives it does the judging. What is undescribed is the process inside extraction. No confidence score is exposed on an extracted figure, no review queue or human verification step is described for ambiguous or non standard disclosures, and nothing states what happens when a company changes how it reports a metric mid history, which is exactly the case where automated extraction silently produces a discontinuity. The source hyperlink lets a user check any figure, but checking is the user's responsibility and no proactive flag on low confidence extraction is described.

The fourth A on this axis and the only one where verification is a property of every individual output rather than of the process around it. Each datapoint is hyperlinked to the filing, footnote, presentation or transcript it came from, so any figure a model produced can be checked in one click against an authoritative document the vendor did not create, which is the Nammu21 property applied universally rather than in principle.

A published benchmark accompanies it, reporting agents reaching approximately 90 percent accuracy on this data against roughly 19 to 20 percent on public web sources, which is a rare quantified comparison in a category where accuracy is usually asserted.

Two qualifications belong on the record and both come from independent commentary: the benchmark does not specify which of three tested agent frameworks produced the full gain, and the claim of ten times the data density of competing providers is vendor reported without independent verification.

More than 160 institutional customers spanning hedge funds, mutual funds and bulge bracket banks, with revenue stated to have more than doubled year on year, roughly 47 million dollars raised and a major investment bank on the register as an investor. Coverage is quantified and checkable at more than 5,500 public companies with thirteen years of history.

The most distinctive customers are not financial institutions at all: three leading artificial intelligence platforms buy the dataset as grounding for their own financial products, which is a form of validation few vendors here can claim, since those buyers evaluate data quality as their core competence. A published benchmark accompanies it, with agents reported to reach approximately 90 percent accuracy on this structured data against roughly 19 to 20 percent on public web sources.

The cross customer training question that dominates this axis is closed by subject matter, since the corpus is public disclosure rather than customer data, so there is nothing belonging to one institution that could improve a model serving a competitor.

On top of that sits a genuine safety architecture: every datapoint carries a hyperlink to the document it was extracted from, which is grounding by construction rather than by instruction, and the company positions the product explicitly against the failure mode of web sourced data producing hallucinations and outputs that cannot be tied back to a trusted source. Held at B because nothing states how customer usage or query data is treated, which is the one stewardship surface that does exist here.

Structurally the cleanest privacy position in this index and it is earned by subject matter rather than by policy. The entire dataset is built from public company disclosures, filings, footnotes, investor presentations and earnings transcripts, so there is no personal data about individuals in the payload at all and no consumer whose records could be exposed. That is a shorter chain even than Nammu21, which carries the same corporate subject property.

The residual question, unaddressed, is customer side rather than subject side: what an analyst queries and models reveals their research direction, which is commercially sensitive and potentially close to material non public information about that firm's intentions, and no statement covers how usage data is handled.

No attestation, certification, trust centre or enumerated framework was located. The consequence of a breach is lower here than for vendors holding consumer or transaction data, since the dataset itself is built from public documents, but institutional buyers of this profile run vendor security reviews regardless, and the newer business of supplying artificial intelligence platforms raises the question of what customer and query data those integrations expose. Nothing published lets a buyer assess it.

No supervisor, statute or instrument is named. That is partly appropriate, since a data provider drawing exclusively on public disclosure carries no licensing obligation of its own and needs none, which is the correct posture under the index convention. What is absent is engagement with the regimes the output enters.

Sell side equity research is a supervised activity with its own record keeping and substantiation requirements, and data feeding a published research note or a valuation used in a transaction sits inside those obligations, yet nothing published addresses how the platform supports them beyond the general claim that traceability meets compliance requirements without extra effort.

No consumer subjects and no protected class analysis applies, so the axis adapts as it did for Hadrius and Federato. The governance question that replaces it is coverage. A universe of 5,500 companies necessarily excludes most listed issuers globally, and extraction quality will not be uniform across it: companies with dense, well structured, English language disclosure are easier to read than smaller, non United States or less conventionally reported ones, which means the issuers already least followed by analysts are the likeliest to be thinly or inaccurately represented in a dataset increasingly used to ground automated research. Nothing published describes which companies are covered, how the universe is selected, or how extraction accuracy varies across it.

No guarantee or indemnity on data accuracy was located, so nothing commercially binds the vendor. What earns the grade is that error correction is built into the product rather than promised alongside it. Because every datapoint links to the document it was extracted from, a user can verify any figure against an authoritative original the vendor did not produce, which makes mistakes findable rather than latent, and that is exactly the property recorded for Nammu21 with the difference that here it applies to every value rather than to a subset.

The limit is that finding an error is not the same as having it fixed: no correction, restatement or notification process is described for a datapoint discovered to be wrong, and nothing addresses what happens to models built on it downstream.

The data chain is fully disclosed by being fully public: filings, footnotes, investor presentations and earnings transcripts are named as the sources and all are documents anyone can obtain, so there is no licensed feed, aggregator or proprietary supplier in the path. That closes the question this axis usually struggles with.

On the model side the protocol layer is described as agnostic and names the frontier platforms it interoperates with, though those are downstream consumers rather than upstream dependencies, and no provider is named for the extraction models themselves. No subprocessor list or hosting arrangement was located.

The machine addressable distribution here is the most complete in this index. A model context protocol server exposes the dataset to language models and agents, it is explicitly model agnostic, and it is integrated with a named financial services product from one frontier laboratory while a custom assistant is published in another's public directory.

Customers are stated to use the data directly inside three major assistant platforms and inside an indexed vendor's own research product, which is the fourth recorded instance of an indexed vendor depending on another. Alongside that sits conventional analyst integration delivering data straight into existing financial models, so the same asset reaches both a human building a spreadsheet and an agent answering a question.

No hosting provider, region selection, residency commitment or private deployment option was located. The exposure is lower than for most vendors here because the underlying data is public disclosure rather than customer records, so what crosses a border is not sensitive in the usual sense. What is not addressed is where customer query and model data rests, which matters for a global institutional client base with its own data governance requirements, and for European users in particular.

No pricing, packaging or basis of charge is published and every route in is a demo request. The question has real shape for this product because there are two very different buyers, an analyst team licensing seats against a company universe and an artificial intelligence platform licensing the dataset for grounding, and nothing indicates whether charging runs per seat, per company covered, per datapoint or per query. Category norm for institutional data, where the incumbents are equally opaque.

The buyer set covers both sides of the market, with buy side hedge funds and mutual funds alongside sell side equity research at bulge bracket banks, plus private equity and valuation teams using the same history for comparable company analysis and corporates using it for their own benchmarking. A newer segment sits outside financial services entirely, selling structured data to artificial intelligence platforms that need verifiable grounding rather than open web retrieval.

What holds this at B is depth rather than breadth of institution type: this is a capital markets research product, and it does not reach banking operations, lending, insurance or wealth advisory, which is a deliberate focus rather than a gap.