How a leading APAC asset manager turned proprietary knowledge into a living asset with Pascal.

A leading APAC asset manager overseeing more than $100 billion had already invested heavily in building its internal knowledge infrastructure. Long before AI entered the picture, the firm had brought together a wealth of proprietary information across its investment organisation, including broker reports, licensed research, internal databases, dashboards, and more.

The problem was that analysts still had to do the work of connecting it.

Answering an internal research question often meant stitching together information scattered across multiple systems and documents. The process became even more tedious because analysts then had to manually compare the information and reconstruct any additional context needed to reach a conclusion.

While the firm had accumulated years of institutional knowledge, accessing that knowledge still depended heavily on an analyst knowing where to look and how the pieces fit together.

When Pascal entered the picture, we began working with the firm's connected internal knowledge systems. Equally important was deploying within their own infrastructure, ensuring the data never left their environment.

So, Pascal became the reasoning layer across the firm's existing information estate. The platform connects to the systems the investment team was already using, so analysts never had to migrate their data into another repository.

What this meant was indexing the firm's documents while preserving the structure and context of each source. Pascal's document intelligence layer maintains the relationship between the information and its original source, so analysts can see where an answer came from and trace it back to the underlying material. A knowledge graph then maps the relationships between the people, companies, documents, events, and other entities contained within that information.

Analysts can now search across their internal corpus directly, ask questions that span multiple sources, and run agentic workflows that retrieve, connect, and synthesise information automatically.

In addition to their internal information, public information is brought into the same workflow. This allows analysts to combine proprietary research with filings, company disclosures, and other external sources without switching between tools.

As the deployment expanded, Pascal moved beyond the firm's core research repositories into email, internal documentation, and other knowledge systems. This opened up a different class of questions.

Answering these questions requires more than finding documents containing the right keywords. The system has to identify the relevant people, companies, documents, events, and dates, then connect them to one another.

A site visit needs to be connected to the analyst who conducted it. That analyst's observations need to be connected to the company discussed. Those observations can then be compared with the investment team's subsequent research and changes in view.

Pascal's knowledge graph provides the relational layer for these connections, while its document intelligence layer preserves the structure and context of the underlying material.

This turns the firm's accumulated research into something analysts can interrogate directly.

The significance of the deployment goes beyond faster search. An investment firm's proprietary knowledge becomes more valuable as it accumulates.

Every research note, management meeting, site visit, analyst observation, and investment decision adds context that can inform future work. Historically, much of that value was locked inside individual documents and systems. Finding it depended on institutional memory, manual search, and an analyst's ability to know which sources to consult.

Pascal adds an intelligence layer across that existing body of knowledge. The firm's research infrastructure remains the source of truth, while Pascal makes the information within it accessible as connected context. As more research is accumulated, that context can be used across more questions, workflows, and investment decisions.

Every answer can be traced back to its underlying sources, giving analysts visibility into where a conclusion came from, whether that was a broker report, internal research note, email, dashboard, or public filing. The architecture also preserves the firm's data sovereignty requirements. Pascal runs within the firm's own infrastructure, allowing proprietary information to remain within the environment controlled by the institution.