Mandate I · Triage
Incoming
Should this opportunity move forward?
Incoming deals, names, and ideas assessed against mandate, policy, and prior work, before they consume an analyst's week. Shape varies by fund. Every team triages.
You know AI matters for your fund. You don't know how to put it to work.
Levercon is the AI implementation team for your fund.
What operators keep telling us.
“We still haven't exactly landed on what we need to do. We just know we need to do some things.”
About
Levercon. Short for leverage consulting.
Your implementation team for AI inside the fund.
Method
How we work. Embed, automate, compound.
i.
Embed with the team
Once we're working together, we sit with your team across the week, not just for an hour. We learn the workflows you wish were faster or more accurate, the checks no-one has time for, and the bits that go wrong when the analyst is on leave.
ii.
Automate the work
Working artefacts in week one, refined every week after. Skills and connected workflows on top of your existing stack. Model-agnostic, no rip-and-replace. We come back, fix what isn't right, and ship the next thing.
iii.
Compound the gains
Every position, every check, every memo adds to the fund's institutional memory. Work that took an analyst a day starts taking an hour, then a minute. The team's reach gets longer without the headcount.
Modes
Where we can help. Two workflows, one system.
Mandate II · Monitoring
Portfolio
Is anything drifting?
Positions, theses, exposures, covenants, and the checks the team hasn't got to yet, flagged before they become problems. Shape varies by fund. Every team monitors.
Founders
The principals.
Questions
Common questions. Direct answers.
Who do you typically work with?
Australian alternative fund managers. Private credit, equities, family office, multi-strategy. Funds with their own investment process and a small enough team that getting the analysts six hours back a week genuinely changes things. Not retail wealth, not US-domiciled funds, not anyone looking for a generic AI consultancy.
What does engagement actually look like? Project or ongoing?
Starts with one hour, free. If it's a fit, we move into a sustained embed: we sit with your team across the week, ship working artefacts, and come back to refine them. It's a relationship, not a project handoff. You can stop at any point and the work stays with you, model-agnostic and inside your tenant.
We're a Microsoft house. Won't IT block this?
We don't tell anyone to leave Microsoft. The patterns we build are model-agnostic. Skills you write today work in Copilot, Claude, or Gemini. Even within Copilot there's usually a 5 to 10x improvement available that the IT team hasn't surfaced.
How does this fit our week? We don't have time.
One hour to start. You show us one thing you do every week that you wish was faster or more accurate. We build a working version live. Worst case, you've spent an hour and seen what AI can actually do. Best case, you walk out with something you can use that afternoon.
Will LLMs train on our data?
Paid tiers from Anthropic, OpenAI, and Google don't train on your data unless you opt in. It's a setting your IT team can lock down. Data stays in your tenant. We don't move it anywhere. The real risks aren't the model provider, they're things like skill files leaking tokens or prompt injection through documents, and those are solvable with engineering hygiene.
How much does this cost?
The first hour is free. After that it depends on what we build together, with a clear scope and price before you commit to anything. We're early enough that we'd rather get the work right than maximise the first invoice.
Recommendation
Start with an hour.
You bring the workflow.
One hour, free. You show us a thing you wish was faster or more accurate. We build a working version while you watch. No deck, no obligation.