“AI is not just a matter of technology but more a matter of change management.” – Arthur Garnier, IT Chief of Staff and Senior Data Scientist at Ardian

Ardian is a global diversified private markets firm. We manage or advise $200bn across Europe, the Americas, Asia and the Middle East, on behalf of governments, institutions, pension funds and high-net-worth investors, providing investment and customized solutions that reflect new economic dynamics and help our clients remain resilient in a changing world.

Recognizing artificial intelligence as a powerful driver of sustainable value creation, Ardian has partnered with Artefact since 2021 to thoroughly harness the potential of AI and generative AI.

This partnership reflects a shared mission to act as architects of lasting change through digital innovation and transformation. While many financial companies struggle to move beyond conceptual pilot programs, Ardian and Artefact have successfully implemented practical AI use cases that deliver measurable and concrete results. One remarkable success story from this ongoing collaboration is Draft AI. Conceived during a collaborative internal company hackathon, this generative AI tool is fundamentally transforming how product teams and investor relations handle complex queries. By prioritizing thoughtful change management alongside tailored technical solutions, the partnership is setting an entirely new standard for AI adoption in private markets.

Challenge: Conquer document overload to rapidly address complex investor inquiries

The private equity industry is inherently document-driven, relying heavily on qualitative reports, financial filings, and lengthy assessments. Investment and product teams constantly face a massive influx of unstructured data, ranging from complex market reports to highly detailed investor questions. Speed and accuracy are paramount factors for success in this space, yet the sheer volume of information creates a significant daily bottleneck for employees.

For Ardian, one major operational hurdle was efficiently managing and responding to extensive investor inquiries. Teams were heavily burdened with finding historical answers buried deep in past communications, which consumed valuable time better spent on high-level strategic analysis. Furthermore, handling highly confidential financial data meant any technological intervention or platform needed to fully comply with strict security protocols and European data regulations.

The true challenge was not simply finding a capable AI model on the market. The real tests were:

  • Identifying the most promising and impactful use cases
  • Building software solutions securely
  • Ensuring teams adopted the tools in their daily routines

The firm needed a highly collaborative approach to cut through the industry hype, empower human analysis, and drastically accelerate response times.

Solution: Empower employees through a grassroots hackathon to design practical AI applications

To address these operational bottlenecks, Ardian and Artefact took a highly collaborative approach to development. The inception of Draft AI began at an internal hackathon specifically designed to generate high-impact use cases across the firm.

Arthur explains the collaborative origins of the initiative: “We organized this hackathon in collaboration with HR, data science, Artefact, and also with the support of Microsoft.” The core idea was to bring together diverse representatives from junior to senior levels across different offices to brainstorm high-impact use cases together.

Artefact played a pivotal role from the earliest stages of ideation through to the final production rollout. Elina Ashkinazi-Ildis, Partner Private Equity at Artefact, recalls the vibrant energy of the initial collaboration: “It was really exciting for the entire Artefact team to be part of this Ardian hackathon,” highlighting that Artefact’s technical teams worked closely alongside Ardian employees, watching as amazing and innovative ideas emerged directly from the staff.

The winning concept was Draft AI, an advanced large language model securely integrated directly into Ardian’s internal database to surface historical data. To bring the functional prototype to life, the joint development team selected an LLM model from OpenAI.

However, Elina emphasizes the complexity of scaling the tool for secure enterprise use: “We picked an LLM from OpenAI and, of course, we built everything in a secure Ardian environment so the team could really try out the solution, see the results, and measure the impact”. She adds that transitioning into full production and deploying the solution across all global teams is a very different ballgame, requiring them to carefully think about the end-to-end process and varied database usage.

Prior to Draft AI, the partnership also successfully deployed GAIA, an internal secure chatbot built around the Azure OpenAI, Claude, and Mistral platforms to handle broader, highly confidential due diligence tasks.

Results: Widespread user adoption and fundamental transformation of daily workflows

The deployment of Draft AI has delivered exceptional outcomes, radically changing how internal teams operate and manage their daily workflows. By matching new investor inquiries with historical answers and seamlessly generating drafted responses, the tool allows employees to handle routine questions quickly and dedicate their focus to highly complex financial issues.

The quantitative impact of the tool is substantial. Arthur highlights the impressive volume of work processed by the new system: “I think that more than 5,000 questions have now been answered thanks to this type of technology.”

Beyond the sheer volume of tasks completed, the most critical marker of long-term success is user adoption. Because the tool was developed collaboratively with direct input from end-users, it integrated seamlessly into existing processes. Today, over half of the product team members utilize Draft AI on a daily basis, representing a massive cultural shift in how work is accomplished.

This high level of engagement did not happen by accident. Elina credits the structural and cultural backing the project received from leadership: “We had huge support from senior management in terms of investment.” She emphasizes that having different members of investor relations teams working with them every step of the way is precisely what made this project such a lasting success.

Looking to the future, the partnership is eager to build upon this incredible momentum. Arthur outlines the bright next steps for the firm: “Our collaboration with Artefact went really great and allowed us to quickly deploy the solution.” He notes that the tool receives incredibly positive feedback, and the firm’s goal is now to actively increase the scope of Draft AI to all Ardian teams interacting with large volumes of requests.