McKinsey, A Century-Old Giant, Embraces Cutting-Edge AI

Mornin’ miners⛏️,

Happy Monday!

Welcome to "Digger Insights" - your daily 5-minute enlightenment on the most recent tech updates! In our enjoyable and easily digestible dispatches, we break down the latest tech trends, giving you a quick and comprehensive view.

Join us, and in just a few minutes each day, you'll gain an advantageous perspective on the ever-evolving tech landscape. Ready to dive in?

Let’s get to it!

Today’s Highlights:

• 🤖McKinsey, A Century-Old Giant, Embraces Cutting-Edge AI⚖️

• 🪙Angeles Equity Partners Raises Over $540M💸

McKinsey, A Century-Old Giant, Embraces Cutting-Edge AI

Generative AI has become a tool essential for plenty of businesses in various sectors. Though some may assume that old companies and sectors tend to stay traditional, this is not always the case. We see more companies every day seemingly staying updated and keeping up with the times, one of them being McKinsey & Company.

McKinsey & Company is one of the largest consulting agencies in the world, first founded in 1926. McKinsey & Company is one of the largest consulting agencies in the world, first founded in 1926. Driven by its confidence in the economic potential generative AI can bring, McKinsey & Company has not only rapidly embraced the use of generative AI, with more than half of its employees utilizing it for their day-to-day workloads, but it has created its own generative AI tool for employees, “Lilli”, despite almost being a whole century old.

The Story of Lilli

“Lilli” is a new chat application designed by McKinsey’s “ClienTech” team, a tech and data team of approximately 200 technologists who power McKinsey’s client-facing tech and data client service capabilities. “Lilli” is named after the first woman McKinsey hired for a professional services role, Lillian Dombrowski, in 1945.

Inspired by Dombrowski’s flexibility, creativity, and thoroughness from back when she worked as the controller and corporate secretary for the firm, everything Dombrowski was is what McKinsey hopes “Lilli” can be for McKinsey employees and colleagues.

Under McKinsey’s chief technology officer (CTO), Jacky Wright, “Lilli” was designed to provide information, insights, data, and plans, as well as recommend internal experts for consulting projects. All the information “Lilli” can provide is based on over 100,000 internal McKinsey documents and interview transcripts, as well as third-party information, all consulted with a network of experts across 70 countries.

According to McKinsey’s senior partner, Erik Roth, Lilli is essentially an AI that is filled to the brim with McKinsey’s knowledge, able to answer every question McKinsey can offer. Roth led the product’s development and has said that about 7,000 McKinsey employees have used Lilli as a “minimum viable product” (MVP).

McKinsey employees’ use of Lilli has lifted a burden off their workloads, creating a more efficient workplace. Research and planning work that would usually take weeks only took hours, and cases that usually require hours to complete only needed minutes while still widening the scope of all inquiries.

In just two weeks, Lilli has already answered 50,000 questions, and 66% percent of McKinsey employees are already utilizing the AI multiple times a week.

Interface and Features

McKinsey’s “Lilli” shares a similar interface to a couple of other public-facing generative AI tools, such as OpenAI’s ChatGPT and Antropic’s Claude 2, with a text entry box for users to enter questions and searches and prompts at the bottom of its window. The responses are generated in chronological chat, as commonly so, showing both the user’s prompts and Lilli’s responses.

Although this is the case, “Lilli” still has its own unique features, one of them being an expandable sidebar on the left with available prompts users can copy, paste, and modify according to their own needs. Roth has said that the list of prompts will also be divided into specific categories soon.

“Lilli” has two different tabs that users can utilize, one being “GenAI Chat,” which sources more generalized LLM backend data, and the other being “Client Capabilities,” trained with more specified data sourced from over 100,000 of McKinsey’s corpus documents, transcripts, and presentations.

This enables “Lilli” users to go through different experiences and compare the two to see which fits them better. Lastly, with the intention to over full attribution to users, “Lilli” provides a whole section of “Sources” below every single response, as well as links and specific page numbers to show where the model drew its response from.

Usage and Ability

McKinsey aims to make “Lilli” a tool for consultants to go through every step of their workload, from gathering research on clients to drafting plans for clients to implement specific tasks and projects.

Roth has stated that the firm plans to experiment further with “Lilli” in this category of work by enabling “Lilli” to securely and privately analyze client information and documentation on McKinsey servers. This feature may take a while to come out into the light, as Roth has asserted that it is still being developed and will not be applicable until it functions perfectly. He believes that “Lilli” has the ability to upload client data safely and securely, as everything that “Lilli” loads faces an extensive compliance risk assessment.

According to McKinsey, “Lilli”’s application was built by themselves, but the firm still utilizes existing LLMs like Cohere and OpenAI to learn more about GenAI Chat and NLP capabilities. The firm views “Lilli” as its own stack, acting as a secure layer between the corpus and the LLMs.

Roth states that McKinsey was “LLM agnostic,” interoperable among various systems, and still exploring new LLMs and AI models depending on which offers the most utility.

Although currently looking to expand usage to McKinsey employees, the firm also has plans to turn “Lilli” into an external-facing product for McKinsey clients or other firms entirely, as McKinsey believes “Lilli” is a tool every organization would need. McKinsey also hasn’t ruled out the option of white-labeling “Lilli.”

Funding News

Angeles Equity Partners LLC Raises Over $540M

Angeles Equity Partners, also known as “Angeles,” is a private equity firm investing in companies of various sectors that could benefit from the firm’s operational transformation and strategic repositioning.

Recently, Angeles closed its Fund II with total capital commitments of over $540 million, comprising institutional investors from North America, Europe, and Asia. The raised funds came from endowments, foundations, public pension plans, fund-of-funds, and family offices.

Angeles will be using the funds collected to invest in underperforming industrial businesses and offer not only investment but also insight and comprehensive operational engagement. Angeles has already committed capital to four platform investments, including Agile Occupational Medicine, Precision Surfacing Solutions, Data Clean, and Custom Goods.

Meme & AI-Generated Picture

Job Posting

• IONQ - Staff Software Engineer, Control - Austin, TX (Remote)

• McKinsey & Company - Sophomore Summer Business Analyst, Internship - North America (In Office)

• Route - GTM Revenue Systems Manager, Salesforce - New York City, NY (Remote/Hybrid)

• Wasabi Technologies - DevSecOps, Infrastructure Team - Boston, MA (Remote)

Promote your product/service to Digger Insights’ Community

Advertise with Digger Insights. Digger Insights’ Miners are professionals and business owners with diverse Industry backgrounds who are looking for interesting and helpful tools, products, services, jobs, events, apps, and books. Email us [email protected]

Your feedback would be greatly appreciated, send it to [email protected]