/ ABOUT

/ HOW IT WORKS

/ APPLICATIONS

/ TEAM

/ RESEARCH

Start the conversation

Two years in stealth

A fundamental AI model for finance

Start the conversation

Two years in stealth

A fundamental AI model for finance

Start the conversation

Scroll down

Read More

0:00 / 0:00

We Can Show You

We Can Show You

See for Yourself

A living world model for fundamental investing

Primordia transforms fragmented financial evidence into a structured, continuously updated model of the investable universe. It connects company fundamentals, management quality, market narratives and macro forces to support rigorous, probability-weighted investment decisions.

Two years in stealth

WHAT PRIMORDIA DOES

Evidence

01

Captures what matters

Financial health, management quality, competitive position, market narratives and macro forces.

Evidence

01

Captures what matters

Financial health, management quality, competitive position, market narratives and macro forces.

World Model

02

Connects the system

Maps companies, sectors, assumptions and causal relationships into one grounded model.

World Model

02

Connects the system

Maps companies, sectors, assumptions and causal relationships into one grounded model.

Reasoning

03

Updates every belief

New evidence propagates through the model, changing assumptions, scenarios and probabilities.

Reasoning

03

Updates every belief

New evidence propagates through the model, changing assumptions, scenarios and probabilities.

Conviction

04

Builds a traceable view

Produces a structured investment thesis with visible risks, assumptions and an auditable reasoning path.

Conviction

04

Builds a traceable view

Produces a structured investment thesis with visible risks, assumptions and an auditable reasoning path.

Applications

One engine. High-stakes decisions

1

Fundamental Equity Research

Understand companies through fundamentals, management, narratives, competitive forces and macro context—not isolated market signals.

2

Wealth Management

3

Institutional Investing

4

High-Stakes Decision Systems

Start Conversation

1

Fundamental Equity Research

Understand companies through fundamentals, management, narratives, competitive forces and macro context—not isolated market signals.

2

Wealth Management

3

Institutional Investing

4

High-Stakes Decision Systems

Start Conversation

Understand companies through fundamentals, management, narratives, competitive forces and macro context—not isolated market signals.
Build more rigorous, explainable investment views across portfolios, sectors and changing market conditions.
Support hedge funds and professional investors with structured reasoning, scenario analysis and traceable conviction.
The same architecture can support complex domains where decisions alter the systems being predicted.

Scientific rigor.

Real capital experience

A team combining neurosymbolic research, institutional investing and hands-on market experience.

Team

Chief Executive

Sandeep Ramesh

Chief Strategy

Thomas Minter

Chief Technology

Harry Stromfelt

Chief AI Science

Rafael Kaufmann

  • Chief Executive

    Sandeep Ramesh

  • Chief Strategy

    Thomas Minter

  • Chief Technology

    Harry Stromfelt

  • Chief AI Science

    Rafael Kaufmann

Reseach paper

Primordia AI Team

Grounded World Models: Efficient and Verifiable Structural Causal Prediction

  • Reseach paper

  • Primordia AI Team

Primordia is a neurosymbolic, multi-agent system designed to augment and automate high-stakes decision-making processes in VUCA environments, with an initial focus on investment analysis. It integrates a symbolic workflow engine with the generative and reasoning capabilities of large language models (LLMs) to produce structured, auditable, and robust analytical outputs. This whitepaper details the system's architecture, the mathe- matical foundations of its components, and the technical implementation that enables its advanced capabilities.

A formal comparison of grounded world models and LLM predictors for investment decisions—showing large gains in cost, latency, and explanation quality through verifiable causal structure.

Read

Draft Technical Whitepaper

A Neurosymbolic Multi-Agent AI For Decision-Making

Primordia AI Team

Under Review for NeSy 2025

The CRISTAL Method: Fast, Reliable Analytical Problem-Solving with Pre-Synthesized Grounded World Models

Author names withheld

Neurosymbolic AI

A Neurosymbolic World Model for Fundamental Investing at Scale

R. Kaufmann, H. Strömfelt, T. Kopinski, G. G. Nicolau, S. Ramesh

We work with a small number of investors, research partners and institutions. Tell us what brought you here.

Create a free website with Framer, the website builder loved by startups, designers and agencies.