ABI NexaIntelligence OS
The data & access layer
ΣFundamentals & Filings DataStandardized and as-reported, citedfxExcel Add-inThe cited number in your model{ }Data APIThe whole corpus, programmatic∞MCP ServerThe AI-native connector⛱Research WorkspaceThe whole file, per name
AI-native deliverables
ΣOperating & Valuation ModelsBuilt, sourced, refreshed on the print▤Company Tearsheets & ProfilesThe as-of-date one-pager✎Confidential Information MemorandaThe sell-side document, drafted✎Investment MemorandaThe buy-side write-up, traceable▣Pitch BooksProfiles and precedents, quoted⚟Screening & Idea GenerationScreens across the corpus
Who we serve
BBuy-SideHedge funds & asset managersSSell-Side ResearchEquity researchIBInvestment Banking & Corp DevM&A and coveragePEPrivate Equity & VCDeal teams, diligence to model▣CorporatesStrategy, IR and FP&A{ }Data & Quant TeamsAgents and pipelines
Plans Methodology Resources Sign in Request access
Products
Fundamentals & Filings DataStandardized and as-reported, citedExcel Add-inThe cited number in your modelData APIThe whole corpus, programmaticMCP ServerThe AI-native connectorResearch WorkspaceThe whole file, per name
Solutions
Operating & Valuation ModelsBuilt, sourced, refreshed on the printCompany Tearsheets & ProfilesThe as-of-date one-pagerConfidential Information MemorandaThe sell-side document, draftedInvestment MemorandaThe buy-side write-up, traceablePitch BooksProfiles and precedents, quotedScreening & Idea GenerationScreens across the corpus
Clients
Buy-SideHedge funds & asset managersSell-Side ResearchEquity researchInvestment Banking & Corp DevM&A and coveragePrivate Equity & VCDeal teams, diligence to modelCorporatesStrategy, IR and FP&AData & Quant TeamsAgents and pipelines
PlansMethodologyResources
Sign inRequest access
Client

An audit-ready layer your agents can prove.

Data and quant teams build on a clean, entitled, cited corpus over an API and an MCP surface, so the output of a pipeline or an agent carries the provenance the business needs.

Request accessSee the platform
What you get

What you actually get to build on.

One corpus over API and MCP

The same provenanced facts every human surface reads, reachable programmatically or through an agent, so you build on one clean layer instead of stitching feeds.

The document layer too

MD&A, press releases and transcripts as structured records, not just the numbers, so text pipelines have a sourced feed to work from.

Every value carries its filing

A number never arrives bare, so whatever your pipeline or agent produces on top can still be traced to source.

Entitled and scoped to your licence

Access resolves to exactly what your organisation licenses, identical across every surface, so nothing leaks through the connector.

One vocabulary across surfaces

The same field catalogue on the API and the add-in, so a convention learned once carries everywhere and mappings stop rotting.

As-reported and standardized, both

Comparable figures and the filer's own captions in one place, so a model chooses the basis it needs.

See it

The same facts, for your agents and your pipelines.

A raw API response, an agent answer and the entitlement rule. Arrows or dots to move through.

One clean layer
Over a REST API
Point your tools at one provenanced corpus over REST, the same facts every human surface reads, with the source filing on every value.
GET /v1/fact
{
  "ticker": "AAPL",
  "field": "revenue",
  "period": "FY2025",
  "value": 416161000000,
  "filing": "10-K",
  "accession": "0000320193-25-000073"
}
Agents that cite
Provenance through the answer
Over the MCP surface a model reasons on the corpus directly and answers with the filing attached, so a pipeline's output proves itself rather than being taken on trust.
MCP
agent> what was AAPL FY2025 gross margin?

tool abinexa.fact // resolved from the entitled corpus

answer> 46.3%, gross profit 192,615 on revenue 416,161.
       cited: 10-K, filed 2025-10-31, accession 0000320193-25-000073
Entitled and scoped
One policy, every surface
Access resolves to exactly what your organisation licenses, identical across every surface, so nothing leaks through the connector.
Entitlementone policy, every surface
RequestResolves to
Licensed market and nameServed, cited
Outside the licenceWithheld, not partial
API, MCP, add-in, workspaceThe same entitlement everywhere
The connector cannot return what the organisation is not licensed for.
In practice

What your pipeline inherits.

01

One clean layer

One provenanced corpus over an API and MCP, the same facts every human surface reads.

cited on every value
By hand today
Stitch together vendor feeds that disagree and cannot cite.
02

Agents that cite

Every response carries the filing, so the agent's output proves itself rather than being trusted.

provenance intact
By hand today
Ship an agent whose answers you cannot audit.
03

One vocabulary

One mnemonic vocabulary across the API and the add-in, so conventions carry over.

shared across surfaces
By hand today
Reconcile field names across every integration.
The surfaces that matter here.
Data API. The corpus, programmatic.MCP Server. The AI-native connector.Fundamentals Data. The cited facts underneath.Research Workspace. The human view of the same data.
Why it fits

Built for a data and quant team.

The citation travels with the fact

A value never arrives bare. The filing rides along with it, so whatever your pipeline builds on top can still point back to the source.

Entitled and scoped

Access resolves to exactly what your organisation licenses, the same on the API and the MCP surface as on every screen a person opens.

The AI layer

The model never does the math.

On ABI Nexa, AI retrieves and explains. It does not invent the numbers, and it cannot pass off a figure it did not fetch.

The math runs in the engine, not the model

Valuations and calculations run in a deterministic engine that reproduces the same result byte for byte. The model narrates that result and checks it for sense; it never produces the number itself.

Every figure resolves to a source

A guardrail checks that each number in an answer traces back to a retrieved cell in the corpus. If it does not resolve, the answer is refused rather than guessed.

Grounded, with the filing attached

Retrieval is scoped to the companies, periods and documents you asked for, and every claim carries the filing, accession and location it came from.

Tested before it ships

A fixed set of questions with known answers runs on every model and prompt change, so a regression is caught by us, not by you.

7,500+
pre-built models, US, India and foreign filers
15 years
of history, restatements kept
29.8M
facts, each one cited
0
wrong on the source-verified sample

See it on your own coverage.

Tell us the names and markets you follow and we will show you the desk on them.

Request accessSign in