01
All your knowledge, connected
We plug into the tools your fund runs on and pull in every document, email and record.
Fund Brain
All your fund's data, in one place.
Instant, cited answers from every deal, doc, email and call your fund has ever produced.
- Data secured
- IT approved
- Recursive self-improvementSelf-improvement
How It Works
From scattered tools to a cited answer.
Four steps turn the data scattered across your fund into answers your team can trust, each traceable to its source.
02
A brain that understands
Every source is connected into one private, permission-aware brain that knows the people, dates and figures in your data.
03
Ask in plain language
Anyone asks a question the way they would ask a colleague, and the brain searches every source for the answer.
04
Take it further
Use the answer as a starting point: run deeper analysis, dive into the data and build on what the brain surfaces.
Usage Dashboard
See exactly how your fund uses it.
Adoption, effort and impact at a glance, so you can double down on what works.
Adoption you can see
Who's using it, how often, and where.
Spot your power users
Your power users, and the habits worth copying.
Find the blind spots
Questions the brain can't answer yet.
Prove the value
Hours it saves the team, every week.
Feedback & Self-Improvement
It learns from every miss.
Every weak answer is captured, grouped and fixed on its own, so the brain never makes the same mistake twice.
Spots the miss
A wrong or thin answer gets flagged, by the brain itself or by you.
Logs it automatically
Captured as feedback in the background, with no forms to fill in.
Finds the pattern
Misses are grouped into themes so the real gap is clear.
Fixes the cause
It learns, so the same mistake does not come back next time.
Continuously, on its own, so it never stops getting better
Data Security
The right people see the right things. Nothing leaves.
Two worries every fund has: the wrong person seeing the wrong data, and something getting out. The brain is built so neither can happen.
Permissions, mirrored
The brain inherits your existing access, person by person. Each person only gets answers from sources they can already see.
Nothing leaves your walls
It lives entirely inside your own environment. Nothing is sent out or exposed beyond the fund.
Never trained on your data
Your documents, questions and answers are never used to train any model, ours included.
Hosted in Australia
Identifiable data stays in Australian regions inside your own tenancy, with no offshore processing.
Pricing & Inclusions
Priced to the build, scoped upfront.
Cost tracks the complexity of your build, its connections, and the state of your data. We scope it with you first, so the number is clear before work begins.
One-off build
$20k–30k+ GST
The final price is set by the complexity of your build, and includes rolling it out across your company.
What the build includes
- A private brain, on your company account
- Every company account set up and ready
- Your tools and sources connected
- Feedback and usage dashboards
- Corrective feedback, captured and applied
- Recursive self-improvement, so it keeps getting better*
Plus ongoing maintenance, improvements, and optimisations, from $200 to $400 per month.Left alone, an AI tool goes stale. We keep yours on a continuous loop of improvement, so it stays sharp and current instead of going static.
“It feels like magic. We ask for something we need and straight away it's cited, sourced, and correct every time.”
Case Study
A private credit fund, with total recall.
We built a Fund Brain for a Melbourne private credit team. Years of memos, models and deal files, now answerable in seconds, down to the figure in the right version of the file.
85,000+
Sources indexed and searchable
25 sec
Average time to answer
260+
Calls to the brain per day
Get Started
Book a strategy call.
See where AI fits in your fund.
We'll show you where AI lifts efficiency and performance across your organisation.
Common questions.
Answers to some frequently asked questions.
Every fund is structured differently. Will it work for the way ours is set up?
It does not assume a standard structure. Instead of changing your data, we build a data scheme: instructions for how your information is organised and where it sits. We have done this across credit and equity funds, and it can answer down to a specific cell on a specific tab, using the latest version even when there are many. It is mapped to your fund, so the variation is the point, not a problem.
Will it respect our access permissions?
Yes. It mirrors your existing permissions. If you can access a source, you will see it when you call the brain; if you cannot, you will not. Nothing changes about who can see what.
What happens when it gets something wrong?
It is captured as feedback, flagged by someone or by the agent itself, rather than left to recur. About twice a week that feedback is grouped into themes, like data that is wrong, stale or missing, and used to improve the agents. You get most of the way there quickly, and the last stretch comes from using it.
Do I get to control what changes, or does it just act on every piece of feedback?
It does not action every comment. It weighs who logged it and whether they own the workflow or just use it, so the owner carries more than a one-off note. Everything runs inside Cowork on your own instructions, so control sits with you.
What if our tools do not have a ready-made connector?
We are not limited to off-the-shelf connectors. When a tool has no native integration, we build custom tools and custom MCP connections so its data still feeds into your brain. If your team relies on it, we find a way to bring it in.
How long does it take to stand up?
A working brain by the end of the first week, then about a month of optimisation as feedback flows through. The exact timeline depends on complexity. We embed on-site for as long as it takes, so you know quickly whether it is valuable.