How Ardabelle became Europe's first AI-Native Private Equity fund

This story was written by Wesype, a Dust Platinum Partner with strong expertise in the financial services industry. It shows how Ardabelle, a new Private Equity firm, became Europe’s first truly AI-native investment organization. With Wesype’s support, Ardabelle has transformed the way PE professionals work.

About Ardabelle

Picture this: It's day one of launching a new Private Equity fund. While focusing on fundraising and deal sourcing, Ardabelle's founders were also asking a different question: what if we could reimagine how investment professionals actually work?

Ardabelle is a Paris-based Private Equity fund founded in 2024 by Virginie Morgon, former CEO of Eurazeo, and six seasoned partners, including Eric Hazan and Julien Gattoni. The fund invests in profitable mid-market companies across industrials, services, and technology sectors, with a clear mission of accelerating the transition toward a sustainable and resilient economy.

But here's what sets them apart: from inception, Ardabelle didn't just adopt AI—they built their entire operating model around it. This wasn't about adding technology to existing processes; it was about fundamentally reimagining how investment teams could transform massive information flows into competitive advantage. The firm's approach reflects a broader shift in the investment landscape, where the ability to process, analyze, and synthesize information faster than competitors has become a critical advantage in deal execution and portfolio value creation.

Challenge: When information overload becomes the enemy of insight

What if the information that should give you an edge was actually slowing you down? Private Equity professionals live in an information-intensive world that demands both surgical precision and lightning speed. But, the more successful you become, the more information you accumulate and the harder it becomes to find what you need when you need it.

Consider a typical Tuesday morning at a traditional PE firm: an analyst receives an urgent request from a partner about market sizing data from a deal worked on six months ago. What follows is a 45-minute archaeological dig through folders, emails, and shared drives ; time that should have been spent on analysis, not information retrieval.

The information archaeology problem

Ardabelle's team overcame the challenges that plague investment professionals everywhere. At a traditional PE firm, analysts spend over two hours daily manually searching through hundreds of deal documents, including Information Memorandums, Vendor Due Diligence reports, and consultant analyses. Each deal generates massive documentation that requires thorough review, but the traditional approach of manual document navigation was consuming valuable analytical bandwidth that should have been focused on insight generation.

Before Dust: the first bricks of a AI-native PE fund

• Ardabelle had already built a lean and modern digital stack, with Notion as the central knowledge hub.
• The team actively monitored 200+ potential investment targets and maintained structured deal documentation.
• Analysts frequently needed to retrieve specific data points, verify sources, and cross-reference information across multiple deals and sectors — a natural challenge given the scale of information handled.
• Knowledge and insights were well captured, but remained tied to individual deal files rather than being instantly reusable across the portfolio.

Julien Gattoni, Founding Partner of Ardabelle — and former Managing Director and CFO of the World Economic Forum — had been an early advocate of Notion, setting it up from day one as Ardabelle’s digital backbone. This gave the firm a strong foundation, but even with this setup, investment professionals inevitably devoted significant time to information retrieval and “reverse lookup” processes — for example, finding the exact source of a market growth assumption.

This was not a weakness of the system, but simply the reality of working with large volumes of data and complex deal documentation.

Active monitoring of more than 200+ companies already gave Ardabelle strong market coverage, but keeping information continuously updated required manual research across multiple sources. This created a natural tension between comprehensive coverage and the time available for deep analytical work.

Investment memos, sector deep dives, and portfolio analyses were already rigorous, but remained time-intensive manual processes that consumed analyst bandwidth. The foundation was solid, but scaling the same quality of analytical output further required an additional layer of leverage.

“We had built a strong digital backbone with Notion from day one, but we quickly saw that our knowledge base could become a bottleneck as we scaled. ”
Julien Gattoni, Founding Partner, Ardabelle

What Ardabelle needed next was not just another tool, but an AI bedrock layered on top of its knowledge base — a way to activate the firm’s data as a living asset. With Dust, Ardabelle’s intent is clear: leverage the knowledge already in place and establish an AI-native foundation — transforming static documentation into dynamic decision support and unlocking scalable, compounding advantages across sourcing, diligence, and portfolio work.

The strategic imperative

In Private Equity, the firm that can analyze opportunities fastest—without sacrificing analytical rigor—wins the best deals. Information archaeology wasn't just inefficient; it was competitively dangerous. Without systematic knowledge management, valuable insights from previous deals, expert calls, and market research remained siloed and difficult to leverage across the broader investment process.

The question wasn't whether to embrace AI. The question was whether to build it internally or find a partner that could deliver enterprise-grade capabilities immediately.

Solution: Building an AI-native investment platform with specialized Dust agents

What if every analyst could access the collective intelligence of your entire deal history in seconds? What if investment memos could write themselves, leaving analysts free to focus on the insights that matter?

Eric Hazan, founding partner at Ardabelle, was a former McKinsey senior partner who had led enterprise AI adoption programs and co-authored the McKinsey Global Institute's "A New Future of Work" (2023) report, understood both the potential and the pitfalls of AI implementation. His research had projected that up to 30% of work hours could be automated by 2030—but only for organizations that approached AI strategically rather than tactically.

With Julien Gattoni, they embarked with Ardabelle’s team on its journey to become the first AI-native PE firm.

Dust won the evaluation

Ardabelle evaluated multiple AI platforms before selecting Dust based on four critical factors:

• Performance flexibility: Access to cutting-edge models from OpenAI and Anthropic through a single platform
• European data sovereignty: GDPR compliance and European hosting for sensitive financial information
• Enterprise security: Built-in confidentiality and permission management for client-sensitive documents
• Integration capabilities: Seamless connectivity with Notion, their centralized knowledge management hub

90-Day AI transformation timeline

The transformation unfolded systematically over three months.

Month 1: Discovery & Design

• Individual interviews with each team member to identify specific pain points
• Use case prioritization based on impact and feasibility
• Initial agent prototypes built and tested

Month 2: Deploy & Iterate

• First production agents launched with immediate productivity gains
• On-demand agent creation process established
• User feedback loops implemented for continuous improvement

Month 3: Scale & Optimize

• Full team adoption achieved with 150+ queries per analyst per week
• Advanced personalization features deployed
• Integration with existing workflows completed

Rather than implementing a broad-based AI transformation program, Ardabelle adopted a targeted, agile approach designed to deliver immediate value while building organizational AI capabilities. The deployment began with individual interviews with each team member to identify their specific pain points, workflow challenges, and high-impact use cases. This personalized approach ensured that AI implementation addressed real operational needs rather than theoretical capabilities.

Instead of developing a comprehensive suite of AI tools upfront, Ardabelle built specialized agents in response to specific use cases, typically delivering functional solutions within days rather than weeks or months. Leveraging their existing Notion-based knowledge management infrastructure, the AI agents were designed to work seamlessly within established workflows, minimizing adoption friction and maximizing immediate utility.

Three specialized agent categories

Ardabelle developed three specialized agent categories, each designed to solve specific operational challenges:

1. The deal intelligence engine: From search to discovery

Think of this as giving each analyst photographic memory of every deal document ever created.

The old way: "I need information about the flexible packaging sector." Cue a 2-hour document hunt.

The Dust way: "What do we know about flexible packaging sector consolidation?" 30-second comprehensive analysis with source citations.

Key capabilities that transformed daily workflows:

• Bidirectional intelligence: Analysts can both discover new insights ("What trends are emerging in sustainable packaging?") and verify existing knowledge ("Where exactly did this 25% margin assumption come from?")
• Cross-deal pattern recognition: The system identifies relevant precedents and comparable analyses from previous transactions, turning historical knowledge into current advantage
• Instant source verification: Every data point comes with exact document and page references, eliminating the credibility questions that plague traditional research

Key insight: This isn't just faster search, it's augmented memory. Analysts now have perfect recall of every insight, assumption, and data point across their firm's entire deal history.

2. The market intelligence autopilot: 200+ targets, zero manual effort

Imagine having a research team that monitors every potential target 24/7 and never takes a day off.

• Before: Ardabelle was already tier-one with Notion databases but still had to do manual tracking, if time allowed for it.
• After: Real-time monitoring with automated briefings and direct database updates

Operational features that redefined market coverage:

• Automated company briefing generation: Standardized profiles incorporating financial performance, strategic developments, and competitive positioning—updated continuously
• Sector deep dive creation: Comprehensive industry analyses that synthesize market trends, competitive dynamics, and investment opportunities without manual research
• Direct database integration: Through the MCP (Model Context Protocol), agents update Notion databases automatically, ensuring the knowledge base stays current without human intervention

Real impact story: When a hot flexible packaging deal came to market, Ardabelle's agents had already identified it as a target six months earlier, tracked its performance metrics, and prepared preliminary analysis. While competitors scrambled to understand the opportunity, Ardabelle was ready to move.

3. The cognitive twin: personalized AI that adapts to you

What if you had an analytical partner who knew exactly how you work, what you prioritize, and how you communicate?

This is perhaps Ardabelle's most innovative implementation: each analyst works alongside a personalized AI agent that adapts to their individual analytical style, communication preferences, and current focus areas.

Personalization in action:

• Adaptive communication: The same analysis request generates different outputs based on the intended audience—sharp efficiency notes for internal use, structured presentations for partners, or tailored communications for portfolio companies
• Individual workflow integration: Each agent learns from the analyst's specific data sources, analytical frameworks, and decision-making patterns
• Context-aware intelligence: Agents maintain awareness of current deals, sector focus, and strategic priorities to provide relevant, timely insights

Results: 5+ hours saved weekly, 30-40% faster execution, and a new competitive edge

Impact metrics: the quantified transformation

Immediate productivity gains that compound daily:

• 5+ hours saved per analyst per week: Time is now spent on analysis rather than information archaeology
• 30-40% faster investment memo production: Tasks that used to take a full day now require 3-4 hours, enabling faster deal execution without compromising analytical rigor
• 150+ AI queries per analyst per week: High utilization demonstrates deep workflow integration rather than occasional usage

Ardabelle can now evaluate 50% more deals in the same timeframe, giving them a significant competitive advantage in hot markets.

Quick win: from information retrieval to insight generation

The cognitive shift that changed everything:

❌ Before AI: Expert insights used once for a single deal 

✅ AI-native Private Equity fund: Expert insights leveraged across the entire deal pipeline

❌ Before AI: New analysts take months to build sector expertise 

✅ AI-native Private Equity fund New analysts access institutional knowledge from day one

Strategic impact: beyond individual productivity

Market differentiation:

• Deeper analytical output: With research automated, analysts spend more time on synthesis and strategic insights
• Enhanced credibility: European data hosting and GDPR compliance provide additional trust for clients concerned about data sovereignty

Organizational capabilities that compound over time:

• Accelerated learning curves: New team members access institutional knowledge immediately rather than building expertise from scratch
• Cross-deal intelligence: Insights and patterns identified in one sector inform investment decisions across the entire portfolio
• Scalable expertise: The firm's analytical capacity grows faster than headcount, enabling selective scaling

Organizational adoption success

The secret wasn't training programs or a mandate from leadership—it was making AI immediately useful for each individual's specific needs:

• Day-one availability: All agents accessible from the start with no barriers to experimentation
• On-demand customization: Ability to create personalized agents within days, not months
• Dedicated support: One full-time resource focused exclusively on adoption and optimization

The result: Instead of forcing people to change their workflows, Dust adapted to how people already worked, then made those workflows dramatically more efficient.

"The transformation isn't just about speed—it's about leveraging our collective intelligence. An insight from an expert call doesn't just inform one deal anymore; it becomes part of our institutional knowledge that benefits every future analysis." Eric Hazan, Founding Partner, Ardabelle

What's next: scaling AI capabilities across the investment lifecycle

This is just the beginning. The question isn't what AI can do for Private Equity—it's what Private Equity will look like when every firm is AI-native.

Immediate expansion plans

Leadership AI adoption: Dedicated workshops for partners and Investment Committee members to fully leverage AI in strategic decision-making processes

Automated LP reporting: AI-powered systems to eliminate manual investor reporting while improving consistency and timeliness

Portfolio-wide intelligence: Cross-portfolio benchmarking to identify best practices, performance patterns, and value creation opportunities at scale

The bigger vision: industrialized investment intelligence

Advanced data integration: Integration with 3PData Snowflake will enable sophisticated analysis of large-volume datasets that exceed current Notion capabilities—think real-time market analysis and competitive intelligence at unprecedented scale

Predictive deal intelligence: AI agents that don't just analyze current opportunities but identify emerging trends and potential targets before they hit the market

Value creation automation: AI-powered systems to support portfolio companies with everything from market research to operational optimization

Every AI capability Ardabelle builds makes the next one more powerful. Their knowledge base becomes richer, their agents become smarter, and their competitive advantage becomes more sustainable.

Your roadmap to becoming AI-native

Based on building Europe's first AI-native PE fund, here's what actually works—and what doesn't.

Start targeted, scale systematically

❌ Don't do this: Try to automate everything at once

✅ Do this: Identify your highest-pain, highest-impact workflows and solve those first

Real example: Ardabelle started with document search, their biggest daily frustration. Success there built confidence and adoption for more sophisticated use cases.

Focus on adoption, not features

The hard truth: Better to have three well-utilized agents than thirty proof-of-concept tools gathering digital dust.

Ardabelle's formula:

• Build agents in days, not months
• Make them immediately useful for individual workflows
• Provide dedicated support for adoption and optimization
• Measure success by usage, not capabilities

Maintain human-AI collaboration principles

The non-negotiable: AI augments judgment, it doesn't replace it.

Practical implementation:

• Always maintain "human in the loop" for strategic decisions
• Design AI to enhance human decision-making, not automate it
• Accept that AI can make mistakes and build verification processes accordingly

"The future of Private Equity isn't about replacing investment professionals—it's about giving them superpowers. AI handles the information processing so humans can focus on the insights, relationships, and judgment that actually create value." Éric Hazan, Founding Partner, Ardabelle

The secret sauce: integration over innovation

The best AI implementation isn't the most sophisticated—it's the one that fits seamlessly into existing workflows while making them dramatically more effective.

Ardabelle's success came from enhancing how their team already worked rather than forcing them to adopt entirely new processes.

The future of Private Equity belongs to firms that can process information faster, analyze opportunities more thoroughly, and execute deals more effectively—while maintaining the human judgment and relationships that remain central to investment success.

The question isn't whether AI will transform Private Equity. The question is whether you'll lead that transformation or be left behind by it.