Investment banking and private equity hire for brilliance, yet they often manage for endurance.
Across the industry, elite teams remain tethered to the manual mechanics of financial reporting and deal screening, tasks that demand hours of precision but offer limited strategic upside. CrossCourtAI is changing that math, automating the "how" of deal-making so teams can double down on the "why."
Sai Naicker knew that rhythm well. Across his career in investment banking and private equity, working on roughly A$2 billion of real asset transactions at firms like J.P. Morgan, Barrenjoey and Cerberus Capital, he saw just how much of the job was consumed by this kind of work. Highly trained deal teams weren’t spending most of their time forming high conviction or sourcing opportunities. They were buried in process; assembling, formatting, and reworking information into the outputs the industry runs on.
It doesn’t make sense.
These are some of the smartest people in finance,” Sai says. “But most of their bandwidth is consumed by an expensive process tax - the manual assembly of information that should be instantaneous.
That frustration stuck with him. And when large language models reached the point where they could reliably generate structured, institutional-grade outputs, not just summaries but fully formatted investment papers, the timing clicked.
What if all of that work could just… happen?
The insight: the work that matters isn’t the work getting done
Inside real estate and private credit firms, the people responsible for making investment decisions are often buried in the work that surrounds those decisions.
Across the industry, around 70% of analyst time is spent on reporting, fund admin and initial deal screening. Tasks that are repetitive, template-driven, and largely manual.
At the same time, the cost of that inefficiency is real. It’s estimated that Australian real estate investors miss around A$1 billion worth of deals each year, not because the opportunities aren’t there, but because teams don’t have the time to properly evaluate them.
The people who should be sourcing deals, running deep diligence, and forming conviction are instead formatting PowerPoints and Excel.
That gap became CrossCourtAI.
Turning hours of work into minutes
CrossCourtAI is built around a system they call Harper: a network of AI agents designed to handle the analytical grunt work of investment teams.
Instead of manually pulling together documents, Harper coordinates the aggregation of deal materials; valuation reports, financial models, diligence packs. It connects directly into the tools teams already use, watching a Dropbox or SharePoint folder, receiving documents over email, or fielding requests through Microsoft Teams. From there, Harper manages a series of specialised agents. One handles market research, another extracts and structures data from documents, and another runs financial modelling.
Within minutes, the system produces an institutional-grade output in the firm’s exact format, from investment committee papers to quarterly reports. Fonts, layouts, structure, all aligned to how that team already works.
Unlike “black box” tools, CrossCourt prioritises auditability and trust. Every line, table and chart is traceable back to its source. Hover over any figure and Harper shows you exactly what was used to produce it, the source reference, the calculation logic and any assumptions.
Harper learns from every iteration, refining its logic based on team feedback to ensure compounding accuracy.
By reclaiming 80% of the time typically lost to screening and reporting, CrossCourt allows firms to evaluate 3x the deal flow and conduct deeper diligence. The result isn't just speed - it’s the ability to run a lean, high-conviction operation with consistency of output.
Three founders, one shared problem
Sai knew the problem intimately, but solving it required more than just firsthand experience.
He spent months looking for the right co-founder before a chance introduction changed things. Sitting in an ice bath in Thailand, talking through the challenge of finding the right person, a friend suggested he reconnect with someone he’d met before; James Oliver.
From their first conversation, it was clear. James had been the founding engineer at V-DAQ, helping scale the company to over A$7 million in ARR. He brought the experience of building and scaling production-grade software, along with deep expertise in AI agent systems.
Randula Ariyawanse joined soon after through mutual connections. With a background as a quantitative engineer at Optiver and experience running an AI consultancy, he added another critical layer; deep technical understanding of both AI systems and financial data pipelines.
Together, the three founders cover the full surface area of the problem: distribution and domain expertise, production-grade engineering, and quantitative AI.
From early conviction to real traction
In the past six months, CrossCourtAI has moved quickly.
The platform now represents around A$7 billion in capital and has processed over A$1bn worth of deals. Their client base includes firms like AsheMorgan, Pindara Partners, Arrowpoint Capital, April Group and Clarke Group, with continued inbound interest from across the market.
One of the key turning points came after winning Startmate’s Pre-Accelerator pitch night in October 2025. That moment created a step change in momentum, leading to new client wins and backing from investors including VentureCrowd, Sydney Angels and Startmate.
Since then, the focus has been on scaling that momentum; bringing larger firms into pilots and converting them into long-term customers.
Building the AI infrastructure for modern deal teams
Today, CrossCourtAI is focused on optimising the most time-consuming parts of deal analysis, fund admin and reporting. But the ambition goes further.
In the near term, the team is working toward onboarding 25+ paying organisations and launching their partnership with Ansarada, giving them direct access to thousands of daily users across real estate workflows.
Longer term, the vision is to evolve Harper into a full dealmaking ecosystem. One that doesn’t just support analysis, but participates in it; from autonomous deal evaluation to AI-powered origination, and even voice-based interaction for senior decision-makers.
The goal is clear: to build the default intelligence layer for investment teams.
A different way to spend time
At its core, CrossCourtAI is built around a simple idea: investment professionals should spend their time on work that requires judgment.
Every hour spent formatting slides or compiling reports is an hour not spent developing insight, building conviction, or winning deals. Over time, those hours compound, not just for individuals, but for entire teams and markets.
CrossCourtAI is betting that by removing that layer of manual work, smaller teams can operate at a much higher level, moving faster, evaluating more opportunities, and ultimately making better decisions.
The ask
If you’re an APAC real estate or private credit investor spending too much time on deal screening and reporting, the team would love to show you what Harper can do.
They’re actively working with firms to run live deals through the platform and demonstrate what this looks like in practice.
Reach out at sai@crosscourtai.com or connect with Sai on LinkedIn.
Watch them pitch at Demo Day
When: Thursday, 30 April @ 7:00 PM (pre-party starting at 5 PM)
Where: Carriageworks (at the close of Blackbird's Sunrise Festival)
What: Pitch night (19 companies)
Tickets: Grab your ticket here
