Brightriver.ai is building the infrastructure for AI-native private markets

Wesley Barnes, co-founder of Brightriver.ai, shares the company’s mission to build foundational AI infrastructure for private markets and why now is the time to act.

“The global demand for private capital is booming, but who’s going to make and manage those investments?” asks Wesley Barnes, co-founder of Brightriver.ai. “[Global] GP and LP assets under management (AUM) are expected to nearly double to $18 trillion by 2030. That’s more than $1.5 trillion in capital needing to be deployed each year, on top of the $2.5 trillion in dry powder already sitting idle. There simply aren’t enough human analysts and seasoned investors to keep pace.”

This imbalance is not just an operational problem. It’s also a structural one. For Barnes, it signals a once-in-a-generation opportunity: “This is why AI is not just a productivity boost. It’s a necessity. The industry’s ability to grow with discipline depends on adopting infrastructure that makes scale possible.”

From automation to infrastructure

Brightriver.ai isn’t just another AI wrapper for investment processes. It’s positioning itself as the ‘AI-native operating system’ for private markets. “We’re building infrastructure, not features,” Barnes emphasises. “Our platform lets clients integrate internal and external data, automate undifferentiated heavy lifting, and drive collaborative analysis across their organisations. We think of it as a single window for integrating data and workflows, purpose-built for private markets.”

Brightriver aims to collapse the distance between unstructured documents, investment theses, and strategic decisions. “From uploading 500-page data-room documents to building dynamic valuation dashboards, we’re helping clients build the connective tissue between their data and their decisions.”

Codifying institutional memory

In a world where basic AI capabilities are becoming commoditised, Barnes sees a key differentiator in how Brightriver enables firms to extract value from their proprietary data.

“Most of the edge in investing comes from judgement and internal pattern recognition,” he explains. “But it’s trapped in people’s heads, or scattered across folders, memos, spreadsheets, and emails. Our platform turns all of that into structured intelligence. We’re not just saving time. We’re helping firms codify their style, their principles, and their memory.”

Brightriver’s search engine builds a knowledge graph from a firm’s historical deals, investment committee memos, CRM notes, and internal reports. This allows users to ask contextually-aware questions and generate recommendations that reflect the firm’s own track record, not just generic industry data available on traditional platforms.

Modular AI agents built for private markets

The platform is built on a modular, agentic AI architecture. Different agents specialise in different tasks: extracting Key Performance Indicators (KPIs) from PDF decks, writing market opportunity memos, benchmarking funds, or parsing redlines in Limited Partnership Agreements (LPAs). These agents can be orchestrated together to match the specific needs of each firm or deal workflow.

“Our clients aren’t using a monolithic chatbot,” Barnes clarifies. “They’re deploying teams of AI analysts that work together. Each agent has access to structured and unstructured data, and the client can always validate, revise, or retrain the outputs.”

Tailored for the last mile of private markets

Private markets remain one of the last bastions of document-heavy, human-driven investing. “These firms don’t operate on clean data feeds,” Barnes notes. “They operate on decks, financial models, data rooms, and relationship insights. You can’t build a real AI solution for private markets unless you start with that.”

Brightriver’s approach has been to build every component of the platform, from its search infrastructure to its Quant and Dataset Agents, with the needs of fund managers and allocators in mind. “We’ve seen teams try to stitch together generic tools like ChatGPT with their own folders and CRMs. It works in isolation, but it doesn’t scale across a team, and it’s not secure or auditable.”

Designed for collaboration and control

One of Brightriver’s guiding principles is that AI outputs should be collaborative and transparent. “There’s a lot of shadow AI happening inside firms,” Barnes warns. “Analysts are using ChatGPT in isolation, copying answers into memos, without oversight. That’s not how institutional processes should work.”

Brightriver’s collaborative workspace architecture ensures that AI-generated insights can be versioned, audited, and improved by the team. It also includes role-based controls and integrates with standard document workflows, helping firms maintain compliance and governance.

Security, privacy, and custom models

Barnes also addresses the concerns many clients have around security and LLM usage. “We currently run eight models, each optimised for different tasks, with zero-retention agreements in place. We’ve also built out a roadmap for on-prem deployments and client-specific fine-tuned models.”

He believes this flexibility is essential for long-term adoption: “You need to meet clients where they are. Some clients are fine with cloud, others need private infrastructure. We support both.”

A growing mandate from LPs and GPs

According to Barnes, interest is accelerating among both LPs and GPs. “We’re seeing LPs start to ask GPs how they plan to use AI, and we’re helping both of them adopt tools that give them better coverage, faster insights, and more robust analysis.”

At the analyst and associate level, Brightriver reduces the burden of information retrieval, cross-referencing, and manual synthesis. That time is reallocated toward higher-order tasks—like speaking with founders, refining investment theses, and building relationships. “We’re not eliminating human input,” Barnes says. “We’re amplifying it.”

The AI-native future of private markets

Looking ahead, Barnes envisions a world where private markets teams operate with the leverage of an entire research department, all powered by AI. “One person will be able to run a $1 billion fund. A five-person team could allocate $100 billion globally. That’s the scale we’re talking about.”

He concludes: “The firms that win in this new world won’t just use AI. They’ll build on it. And we’re giving them the foundation to do that.”