A European Mid-Cap PE Firm

How a European mid-cap private equity firm built custom Crunched agents to compress days of company research into hours — and ship every IC pack with full confidence.

Core metrics

• 60%+ time saved on end-to-end company screening
• 3× more cases taken to full op-model depth
• 100% IC packs auto-reviewed before partner sign-off

About the client

Client: Anonymized — European mid-cap PE firm
Industry: Private Equity (Mid-Cap, European)
Primary Use Case: Investment screening, operating model builds, IC pack production, model and deck QC
Product: Crunched — custom Excel + PowerPoint agents

"We're not only buying time savings — we're buying confidence. When an IC deck goes out with a partner's name on it, knowing every number reconciles end-to-end is what we're paying for."
— Investment Director, European mid-cap PE firm

Executive Summary

European mid-cap private equity is one of the most throughput-intensive segments in the asset class. A single deal team screens dozens of companies a quarter against compressed timelines, and most candidates never get the operating-model treatment they deserve simply because the cost of building one is too high relative to the conviction available at that stage.

This firm took an unusually forward-leaning view: instead of treating AI as a chat sidebar, they partnered with Crunched to build firm-specific agents that own end-to-end workflows — from market research, to populating their standard operating model from messy primary sources, to producing the first three slides of a company intro in their own deck format. The same agents are now used to error-check counterparty models and IC decks for cross-document consistency.

The operational change is substantial. Initial company analyses that used to take several hours now take under an hour. The team takes more cases to full operating-model depth — including base, high, and low scenarios — that previously would have been screened out for time. And every IC pack runs through Crunched before it goes near a partner.

The Challenge

Throughput, not talent, is the bottleneck

Mid-cap PE teams are perpetually triaging. Most opportunities arrive as a name, a sector, and a thesis fragment — and the team has to decide, fast, whether it's worth more time. In the Nordics, that triage involves financials buried in Proff and other registries; in the rest of Europe, it's annual report PDFs and patchy market data. Either way, the firm-specific operating model needs to be populated cleanly and consistently for any cross-target comparison to mean anything.

The work isn't intellectually hard. It's slow, manual, and unforgiving — wrong assumptions compound, sources don't reconcile, and any error caught late means re-running the model. In practice, it means most names get a directional memo and a quick comparable, and only a fraction get the operating-model treatment they deserve. Decisions about which deals to chase end up being made on shallow analysis.

The firm's leadership saw the bottleneck clearly: it wasn't analytical talent, it was throughput. They wanted associates and directors spending less time copying data into models, and more time building investment views.

The Solution

Custom agents, not a chat sidebar

Rather than adopt a generic AI tool, the firm partnered with Crunched to build a layered set of custom agents tuned to their specific workflow. Two of those agents anchor the day-to-day.

1. The Company Screening Agent
   The flagship agent compresses the firm's entire initial company review into a single run. It pulls company and market research, sources financials directly from Proff for Nordic targets and from annual-report PDFs for continental ones, builds a populated three-statement operating model in the firm's standard structure — with the option to specify base / high / low cases or a more granular operating build — and outputs the first three slides in the firm's deck format, ready for the analyst to pressure-test. What used to take several hours for a junior team is now under an hour.

2. The Customer Cube Agent
   The customer cube agent applies the firm's own NRR and churn definitions and segmentation to any target's cohort data, producing a clean retention view from a messy export. The output is standardized — same definitions, same cuts, every time — so two analysts looking at two different deals are reading the same shape of analysis. Hours of cleanup compress into a templated output the team can trust.

3. Model and deck QC
   Around those two agents sit two additional safeguards. Every Excel model received from a counterparty — and every model the firm builds itself — runs through Crunched's error-check, catching broken links, calculation-flow problems, and the categorical miss (a missing D&A add-back, a sum that drops a row when a tab moves) that doesn't surface natively in Excel. PowerPoint cross-deck error-checking is in early access, with the firm one of the first design partners — the agent looks across exec summary and downstream slides for inconsistencies before anything goes to IC.

"Excel models from counterparties always come in with something hidden — a reference that breaks if you change a tab, a sum that misses a line. Catching that on Friday night used to be the job. Now it's a 5-minute check before I open the model myself."
— Investment Director, European mid-cap PE firm

Focus Area

Where Crunched plugs into the workflow

The Results

Faster screening, deeper diligence, fewer late-stage surprises

Quantifiable outcomes

• 60%+ time saved on the end-to-end initial company screening process — research, operating model, and first IC slides.
• 3× more cases taken to full operating-model depth before screening decisions, meaning fewer good companies fall through on shallow analysis.
• 100% of IC packs and counterparty models auto-reviewed before going to a partner, materially reducing the cost of late-stage error-catching.

Qualitative impact

• Confidence: the team describes the value not as "time saved" but as "knowing the deck is right before it leaves my desk."
• Culture: the firm has gone from no AI tools at the desk to a shared, standardized set of firm-specific agents — one of the more forward-leaning AI postures in European mid-cap PE.
• Capacity: associates and directors are spending more of their time on investment views and less on data wrangling.

"Before, we'd build out a real op model on maybe one or two names from a screening batch. Now we build them on five — and the bar for what makes it past initial review is much higher because the analysis is much deeper."
— Senior Associate, European mid-cap PE firm

Broader Impact

What AI-native mid-cap PE looks like

European mid-cap private equity is a segment defined by deal volume and decision speed. The firms that win are the ones that look at more companies, more deeply, in less time — without compromising rigor at IC. By building custom Crunched agents around their own data sources, model templates, and deck formats, this firm is doing exactly that. They are an early example of what AI-native mid-cap PE looks like: not a chat tool bolted onto Excel, but firm-specific intelligence that takes ownership of the slow, deterministic work, giving partners and associates back the hours they need to actually think about deals.

"If a peer fund is still doing initial company analysis the old way in 18 months, that's a competitive gap. We see this as core infrastructure now."
— Investment Director, European mid-cap PE firm

About Crunched

Crunched is an AI Excel analyst built by and for Excel power users in investment banking, private equity, and management consulting. Backed by First Round Capital, Y Combinator, 20VC and others, Crunched integrates advanced AI models directly into Excel to automate modeling, error-checking, research, and templated outputs — without breaking existing workflows.

The goal: automate Excel grunt work, one high-impact pain point at a time.