ASCEND
/

The financial AI infrastructure
we invest with every day

ASCEND invests its own capital across global markets on this infrastructure. Every product we offer, and every system we build for institutions, runs on it too.

1,217+
Instruments traded
$2B+
Cumulative executed notional
750+
Connected data nodes
1M+
Graph and scenario evaluations per day
  1. ApplicationsOur products and your own
  2. LLM and agentsRole-separated models and checks
  3. OntologyEvery tradable asset and its links
  4. Data pipelinePrices, filings, news, macro, your data
Path by product

01 Data pipeline

Market data from everywhere,
stamped with when it was known.

Everything collected is normalised and versioned.

  • Market dataPrices, volumes and curves from exchanges worldwide
  • Filings and regulationDisclosures, prospectuses, rule changes
  • News and policyNewswires, central banks, government announcements
  • Macro indicatorsRates, FX, trade and economic releases
  • Your recordsHoldings, contracts, models and documents you connect

Rules

Point-in-time
Every figure carries the time it was known, and corrections are kept as new versions.
Numbers from code
Every figure is calculated in code, not written by the model.
Primary sources first
Filings and contracts come before data vendors and news.

02 Ontology

Every fact,
linked in one graph.

One event’s impact is calculated across every linked entity, asset, product and portfolio.

EventOvernight chip export curbs

Entity and instrumentListed shares of the companies named

  1. Held directly30%
  2. Chip ETF15%
  3. 2x ETF, levered10%
  4. Inside funds4.5%

PortfolioClient’s real chip exposure

59.5%

EventOvernight chip export curbs

Entity and instrumentListed shares of the companies named

  1. Held directly30%
  2. Chip ETF15%
  3. 2x ETF, levered10%
  4. Inside funds4.5%

PortfolioClient’s real chip exposure59.5%

03 LLM and agents

Only verified output
reaches your client.

Analysis and verification are done by separate agents.

Agent roles

  1. 01Planning
  2. 02Financial analysis
  3. 03Causal analysis
  4. 04Challenge
  5. 05Evidence check
  6. 06Number check
  7. 07Drafting

Checks

  • DataReference date, revisions, usage rights
  • EvidenceThe source supports the claim
  • NumbersUnit, currency, period, formula
  • CausalityMechanism and assumptions
  • ResearchCompeting views and missing data
  • ComplianceWording, suitability, channel policy

04 Applications

Every product runs
on one infrastructure.

Dedicated systems,
built on this infrastructure.

Build a dedicated systemKeep the data and checks we invest with; only the screens and deployment are built around you.Adopt a solutionProducts for commodities, banks, wealth, PE and VC, connected to your data and approvals.