Investment team
Deal screening & scoring
CIMs and targets scored against your own fund's thesis — red flags, EBITDA adjustments and concentration surfaced before the first meeting.
AI Agents
Useful AI starts with structured data.
The Helm structures your fund's data — numbers reconciled, one set of definitions. That structure is what turns AI from a conference slide into a working tool: ask in plain language today, agents built on the fund's experience next, bespoke modules where you need them.
Three degrees of AI use.
Structured data is the entry ticket.
Ask a question, get the number — cited, sourced, in seconds. Behind that simplicity sits a data model that reconciles your figures and opens them to AI: yours or ours.
Ask your portfolio. In your words.
YouHow has Portco A performed over the last three years?
Reading the portfolio…
Revenue and EBITDA have compounded ahead of the invest case since entry, with EBITDA running €2.3m above plan this year. The position marks at 2.15× with the exit slated for Q2 2027.
2.15×Gross MOC
28.5%IRR
+€2.3mEBITDA vs plan
Fund VII · Portco A · reconciled actuals to Jun 2026
YouWhat happens to fund IRR if we push the Portco A exit to 2028?
Reading the portfolio…
Sliding the exit from Q2 2027 to mid-2028 holds the target multiple but stretches the hold to 8 years. Fund VII gross IRR moves from 22.4% to 20.6% unless exit EBITDA improves by ~€4m to compensate.
−1.8ptFund gross IRR
3.06×Fund MOC, unchanged
+€4mEBITDA to offset
Fund VII · scenario vs current exit plan
YouWhere does the Portco B value-creation plan stand?
Reading the portfolio…
The plan is 62% complete. Procurement and the pricing rollout are on track; the working-capital workstream is two quarters behind. The remaining initiatives carry an estimated €9m of EBITDA still to capture before exit.
62%VCP progress
1 of 6Workstreams behind
€9mEBITDA remaining
Fund VII · Portco B · VCP tracker, Q2 2026
YouHow has Portco C's EBITDA moved over the last three months?
Reading the portfolio…
EBITDA stepped from €3.8m to €4.4m over the quarter, with margin at 18.2% — 0.6pt above plan. The bridge attributes most of the gain to the April price increase holding without volume loss.
€4.4mEBITDA, latest month
18.2%Margin
+0.6ptvs plan
Fund VIII · Portco C · monthly actuals
- Your AI tools
- Plug Claude, ChatGPT or Mistral into your portfolio over MCP — your subscriptions, your accounts, your rules.
- Or inside The Helm
- The same layer runs in the product, for teams that keep everything in one place.
- Answers that cite
- Every figure traces back to the reconciled data — no black-box numbers.
Never pay twice for the same advice.
Ten years of external missions — pricing, diligence, 100-day plans — indexed into the fund's knowledge base. Every new mission compounds it, your agents retrieve from it. Analysis first; the decision stays yours.
Across the value-creation plan
Top line & pricingOperational efficiencyDigital & AICash & working capitalM&A & build-upExit preparation
Let's build your bespoke AI use cases, made for private equity.
Investment team
IC memo drafting
The first draft in your committee's format, assembled from the deal's documents and your past deals — the team edits instead of compiling.
Portfolio operations
Getting a company data-ready
Structure the reporting of a portco still on spreadsheets: collection, normalisation, definitions — the company becomes steerable and comparable to the others.
Portfolio operations
Bespoke operating cockpit
A company's own trade KPIs — unit economics, like-for-like, service levels — put under control, where standard monitoring stops.
Fund Services
LP reporting
Your quarterly reports in house format, assembled from reconciled data and tailored per investor — the manual assembly disappears.
And yours?
This list stops where your fund begins. Describe the process — we'll tell you whether it holds as a module.
Any recurring process built on fund data can become a module.
What funds ask us.
Which AI models do you use?
The ones you choose. Over MCP, the conversational layer plugs into your own Claude, ChatGPT or Mistral accounts. Inside The Helm, the model is agreed at setup, like the rest of the deployment — nothing is imposed.
Does our data train a model?
No. Your portfolio data never trains a model — not ours, not a provider's.
Where is our data hosted?
In the European Union, and it doesn't leave it. The conversational layer runs through AI accounts you already control — or through The Helm; nothing is routed to a third party you haven't chosen.
Who sees what?
Each fund's data model is isolated. An agent working for your fund sees your fund — nothing else.
Do we have to change our tools?
No. The conversational layer comes to where you already work — your AI tools connect to your portfolio, not the other way round. The Helm remains the reference surface for the numbers.
What is actually live today?
Rung one — the conversational layer — runs on the platform today. Fund agents are in deployment: first agents exist, the rollout is underway. Custom modules are scoped case by case, as a project with your team.
The direction
The means of the largest asset managers, at the scale of your fund.
Thirty minutes to understand your fund and what it needs, and prepare a tailored demonstration — on your own data, if you want it.