The Pascal platform

The agentic research platform for institutional investors.

Pascal plugs into Bloomberg, FactSet, your RMS, your data lake, and your LLM of choice. Deploy the full platform or layer in the components you need.

Who this is for

Built for every stakeholder in the investment process.

Portfolio Manager

Run your morning pack before markets open.

Overnight macro, earnings surprises, and portfolio alerts — assembled automatically and cited to the source. Your conviction call, not someone else's summary.

Morning PacksCatalyst AlertsPortfolio Monitor
Research Analyst

First-draft IC memo in the time it takes to read a filing.

Pascal reads the 10-K, the transcript, four broker notes, and your internal thesis — and returns a structured memo with every claim pinned to a source line.

Earnings SynthesisIC MemosCompany Primers
CTO / Head of Technology

Deploy enterprise AI without touching your data perimeter.

On-prem or VPC-native. SOC 2 certified. Model-agnostic. Pascal plugs into Bloomberg, Snowflake, and your RMS without requiring a single API key to leave your network.

SOC 2 Type IIVPC DeploymentMCP Native
Chief Compliance Officer

Every output is auditable, attributed, and regulator-ready.

Full citation trails on every research output. RBAC enforced at the infrastructure layer. Audit logs exportable for SEC, FINRA, and MRM review — with zero manual effort.

Audit TrailRBACSource Citations
Agent Control Center

One platform. Every researcher.

One orchestration layer for every agent, every query, every data flow — all inside your perimeter.

01

Manage Sources and Permissions

Scope every query by company, document type, integration, or private vault. RBAC and audit logs enforced on every action.

Select Source

Goldman SachsSELECTED

Document Types

📁

My Vault

Drop files or browse to upload

02

Secure Workflow Routing

Queries move between agents, sources, and models without leaving your perimeter. Audit logs and citations are built in.

LLM Layer

Opus 4.7
GPT-5.5
Gemini 3.1

Pascal Cognitive Engine

Routing · Audit · Citations · Governance

Finance Data Layer

BloombergFactSetBroker ResearchInternal DocsRMS
03

Choose Any Model

Route research to Claude, GPT, or Gemini per task. Auto Mode picks the right model when you don't.

Model Selection

Auto Mode

Auto Mode

Optimal model selected per task

Opus-4.7

Long-form reasoning & analysis

GPT-5.5

Fast synthesis & summarisation

Gemini-3.1

Multimodal & structured inputs

Finance Ready MCP Connectors

The data stack your investment team already pays for, connected.

Connectors built for institutional finance data — Bloomberg, FactSet, SharePoint, Snowflake, your CRM. Not retrofitted from consumer tools.

01

Ready Integrations Out of the Box

SharePoint, OneDrive, Google Drive, Box, Snowflake, Databricks, DealCloud, Notion — connected out of the box. No bespoke pipelines.

Available Connections

SharePoint

Connected ✓

OneDrive

Connected ✓

Google Drive

Connected ✓

Box

Connected ✓

Snowflake

Connected ✓

Databricks

Connected ✓

Notion

Connected ✓

DC

DealCloud

Connected ✓

02

Purpose-Built for Financial Workflows

Built for the workflows investment teams actually run: KYC, source-of-wealth, due diligence, portfolio updates, earnings, IC memo prep.

KYC Dossier — Blackstone Group

Generated · Nov 14, 2025 · 3 sources

Export ↑
BloombergSEC FilingInternal Notes
Entity Overviewsource ↑

Blackstone Group Inc. is a leading global alternative asset management firm with $1T+ in AUM across PE, RE, credit, and hedge funds.

Key Shareholderssource ↑

22.4% — Blackstone Management Partners LLC

Business Segments

Private Equity, Real Estate, Hedge Fund Solutions, Credit & Insurance

✓ Source Cited✓ Audit Trail✓ Compliance Ready
Financial Agents

25+ research agents purpose-built for institutional investors.

Pre-built agents, composable workflows, real-time catalyst tracking. All inside your governance perimeter.

01

Run Pre-Built Investment Agents

KPI extraction, research memos on your house template, generative matrices across company sets. Live on day one.

All AgentsMy AgentsShared
📊

Generative Matrix

Cross-company KPI extraction

Run
🔍

KPI Extractor

Filing-level financial metrics

Run
📄

Research Memo Generator

Analyst-grade memo drafting

Run
📈

Bull-Bear Analyzer

Thesis generation & scoring

Run
02

Compose Workflows for Every Research Cadence

Earnings Summary, IPO Prospectus Analysis, Bull-Bear, Company Research Pack, KYC Dossier. Run them as-is or compose your own.

AllEarningsResearchDue Diligence
📋

Earnings Summary

~2 min

Run
🏦

IPO Prospectus Analysis

~4 min

Run
⚖️

Bull-Bear Analysis

~3 min

Run
🔎

Company Research Pack

~5 min

Run

KYC Dossier

~6 min

Run
03

Stay Ahead of Every Catalyst

Pascal's calendar auto-fetches filings, transcripts, and investor decks the moment they drop.

Earnings Calendar

Nov 2025 · 4 events

METANASDAQ

Meta Platforms Inc.

Q3 Results

Nov 12

NVDANASDAQ

Nvidia Corporation

Q3 Results

Nov 13

JPMNYSE

JPMorgan Chase & Co.

Q3 Results

Nov 14

Today

VNYSE

Visa Inc.

Q3 Results

Nov 15

Context & Memory Graph

Your research compounds. Every query gets smarter.

Pascal builds a persistent knowledge graph across every document, filing, and output your team has ever touched.

01

Persistent Entity Memory

Every company, sector, and theme gets a living memory layer. Past memos, alerts, filings, and notes are linked and retrievable on demand.

Entity Memory — Nvidia Corp.

Nov 14

Memo

Q3 Earnings Analysis

Source: 4 sources

Oct 28

Alert

Supply chain risk flagged

Source: Bloomberg

Oct 10

Filing

10-Q ingested & parsed

Source: SEC EDGAR

Sep 22

Note

Broker report summarized

Source: GS Research

02

Cross-Session Context

Analysts pick up exactly where they left off. Pascal surfaces prior theses, flagged risks, and related documents automatically — without re-prompting.

Research Memory Graph

Research MemoFilingDocumentRisk Alert
03

Institutional Knowledge Capture

Every output, annotation, and analyst decision is stored and attributed. The firm's analytical edge compounds with every deal and every quarter.

Entity Memory — Nvidia Corp.

Nov 14

Memo

Q3 Earnings Analysis

Source: 4 sources

Oct 28

Alert

Supply chain risk flagged

Source: Bloomberg

Oct 10

Filing

10-Q ingested & parsed

Source: SEC EDGAR

Sep 22

Note

Broker report summarized

Source: GS Research

Pascal AI Front End

The workspace where analysts run every Pascal agent.

From raw filings to polished IC memos. Every output reshapes to the workflow and traces back to source.

01

Chat With Any File or Data Source

Filings, transcripts, broker reports, internal memos. Indexed alongside live data feeds, every answer cited.

📂

Drop files or click to upload

PDFs, Excel, Word, CSV — any format

Ask anything about your documents...
02

Adapt the View to the Workflow

The same query becomes a KYC dossier, an IC memo, or an agentic table. Pascal reshapes the output to the task.

Company Primer — Marriott International

TableTextExport ↑

Business Overview

Marriott International is the world's largest hotel company, operating 30+ brands and 8,800+ properties globally. Asset-light franchise model generates strong free cash flow with global RevPAR exposure.

Financial Highlights

Revenue

$23.7B

EBITDA

$4.1B

FCF

$3.2B

Key Risks

Macro sensitivity to travel demand · FX exposure on international properties · Capex for new property development pipeline

Sources:SEC FilingAnnual Report 2024Broker Notes
03

Analyze Data with the Agentic Table

Add columns — ROE, revenue growth, CET1, cost-to-income. Pascal extracts each value from filings, with the citation underneath.

Agentic Table — Banking Sector

+ Add Column
CompanyROERev Growth YoYCost-to-Income
JPMorgan Chase17.2%+8.4%55.2%
Goldman Sachs13.8%+11.2%62.1%
Wells Fargo11.4%+5.7%63.8%
Citigroup8.9%+3.2%67.4%
All values sourced from10-K filingsEarnings releases
For Developers & Engineers

API, SDK, and Toolkit

Integrate Pascal into your internal systems, compose custom agents, and automate end-to-end research workflows — in Python or TypeScript.

01

Embed AI Into Internal Systems

Plug Pascal directly into the systems your investment team already uses — RMS, OMS, internal portals, research dashboards, and CRMs.

pascal_sdk.py
Python
1from pascal import Pascal
2import os
 
4# Initialize Pascal client
5pascal = Pascal(
6 org_id="org_5f2b1c9a_demo"
7 api_token=os.getenv("PASCAL_TOKEN")
8 instance_id="instance-usr-001"
9)
 
11# Query your RMS portfolio
12portfolio = pascal.query.run(
13 source="rms_production"
14 company="Nvidia Corporation"
15 type="portfolio_exposure"
16)
02

Build Custom Logic With the SDK

Pascal's Python and TypeScript SDKs let your engineering team compose custom agents, chain workflows, and build firm-specific logic on top of the Cognitive Engine.

pascal_sdk.py
Python
1from pascal.sdk import AgentChain, Workflow
 
3# Compose a multi-step workflow
4chain = AgentChain([
5 AgentChain.step("kpi-extractor", {
6 "tickers": ["NVDA", "MSFT"],
7 "metrics": ["ROE", "EBITDA"],
8 }),
9 AgentChain.step("research-memo", {
10 "template": "ic_template_v2",
11 "audience": "Investment Committee",
12 }),
13])
 
15result = pascal.workflows.run(chain)
16print(result.memo, result.citations)
03

Automate End-to-End Processes

Run Pascal agents headlessly through the API to power scheduled and event-triggered workflows — pre-market earnings packs at 6 a.m., portfolio alerts on regulatory filings, IC memo first drafts.

pascal_sdk.py
Python
1from pascal import Pascal
2import os
 
4# Initialize Pascal client
5pascal = Pascal(
6 org_id="org_5f2b1c9a_demo",
7 api_token="tok_demo_pk_abc123DEF456ghi789",
8 instance_id="instance-usr-001",
9 workspace_key="ws_sk_test_01a2b3c4"
10)
 
12# Run Earnings Summary Agent
13response = pascal.agents.run({
14 "agentId": "earnings-summary",
15 "input": {
16 "objective": "Generate Q2 earnings summary for US mega-cap tech",
17 "tickers": ["NVDA", "MSFT", "AAPL"],
18 "audience": "Portfolio Manager"
19 }
20})
21print(response)
Trusted partners & infrastructure

Backed by leading capital allocatorsbuilt on best-in-class AI infrastructure

Built for your perimeter

Flexible Deployment, Sovereign Control

Pascal runs as a managed service, inside your private cloud, or fully on-premise, with single-tenant isolation available. Your privileged data stays inside your perimeter and never trains a shared model.

  • AES-256 encryption at rest and in transit
  • Zero data retention — outputs never persisted
  • Role-based access control with full audit trail
  • SOC 2 Type II · ISO 27001 certified

Pascal Cloud

Managed service

  • Auto-scaling
  • SOC 2 verified
  • Zero-config deploy

Private Cloud

VPC / Private

  • Your cloud env
  • Customer-managed keys
  • Single-tenant

On-Premise

Full control

  • Air-gapped option
  • On-prem compute
  • Custom deploy

Single Tenant

Maximum isolation

  • Dedicated infra
  • Zero data sharing
  • Dedicated SLA
Trust & Compliance

Security cleared for institutional finance.

SOC 2 Type II, ISO 27001, GDPR.
Audited by independent third parties. Built for the firms that get reviewed every quarter.

SOC 2

Type II

ISO 27001

Security

GDPR

Coming Soon

USA

228 Park Ave S PMB 55152, New York

UK

86-90 Paul Street, EC2A 4NE, London

India

Indiqube Edge, ORR, Bangalore

Copyright © 2026 | All Rights Reserved

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.