Boston Consulting Group is “all in” when it comes to developing AI tools, Scott Wilder, a Dallas-based partner at the firm, told Business Insider.
About 15 months ago, the company launched an in-house research and development lab. This is one of several ways we are developing new AI-based technologies to accelerate our operations.
“Every company has to become a technology company,” he said. “BCG is no exception.”
Since OpenAI introduced ChatGPT in 2022, generative AI has become a shiny object that enterprises can't afford to ignore. And consulting firms from McKinsey to BCG, once known for strategy and advisory services, are shifting more toward building, implementing, and maintaining AI tools for their clients.
“As a computer scientist, it's interesting to see service companies change like this,” said Wilder, who studied computer science at the University of Texas at Austin.
“Forward deployment”
At BCG, Wilder said generative AI innovation is currently happening at three levels.
First, there is the data level, which is driven by data engineers. “We're at the point where we're really starting to leverage our internal proprietary data,” Wilder said.
Wilder said BCG is building MCP servers and agents on top of select internal public data sources, allowing AI tools to automatically retrieve the right information at the right time. He sees this as a shift in the future of AI, where model intelligence becomes increasingly tied to knowing what data to use and when.
Then there's a “middle layer” of innovation that comes from consultants building tools and agents for their clients' projects, Wilder said. When one of these tools is a hit, it goes back to the R&D team and integrates the tool into a central marketplace that can be used by the rest of the company.
Finally, there are top executives who are developing company-wide tools.
These include Deckster, a slideshow editor trained with 800-900 slide templates that allows consultants to quickly create presentations; Deckster also has its own development team. We also have Ava, an internal help tool for IT issues, HR questions, and other internal requests. There's also GENE, a conversational chatbot that the company uses for presentations, podcasts, and promotional efforts.
Wilder said the company has about seven or eight top-level tools.
Wilder said the company also has a team of “forward deployment consultants.” The role is inspired by the forward-located engineers popularized at Palantir, where he builds vibe coding and tools for clients.
“We put the best in the world into the case, and the case makes it reusable. It goes back to the research and development team,” he said. “So we're seeing a lot of innovation happening from there, and a lot of that needs to happen on the rock wall and on the incident team, not in the middle.”
These consultants combine external coding platforms and internal resources to build tools like OpenAI's ChatGPT. BCG bills itself as the No. 1 creator of custom GPTs worldwide, having created about 36,000 to date, Wilder said.
new assembly line
At the same time, the company's new research and development team is working to disseminate these innovations within the company.
When the R&D team was launched about 15 months ago, its mission was to build a knowledge graph from the company's slide decks, transcripts, and other internal documents. Knowledge graphs are commonly used to organize and connect data across multiple sources. So, for example, from a meeting transcript, the research and development team will aim to extract not just words but also concepts and emotions, Wilder said.
Currently, the team is working on fleshing out the ideas generated by consultants working with clients on the front lines.
Wilder said the new tools will also be inspected by BCG's Red Team team, which simulates adversarial exercises to address security vulnerabilities in the tools. It is also reviewed by the Data Protection Authority to ensure compliance with data privacy laws, undergoes legal review, and is assessed for information security.
Once ready, it will be passed to the company's Marketplace team and integrated into a central repository of tools that can be used by the rest of the company, Wilder said. “I actually have an agent,” he said. “This orchestration agent basically helps people understand what are the best tools and data to use within BCG.”
Wilder said the company prides itself on its “bottom-up” approach to innovation. “Probably 80 percent of our custom GPT comes from the front lines,” he said.
New ideas also surface through the company's “realization network.” This is a team of about 1,000 people from all departments, from human resources to legal, who act as learning and development coordinators and train the rest of the company on AI.
Wilder said they also serve as “our eyes and ears.”
The team meets monthly to share ideas and insights gleaned from working with different teams.
Wilder said BCG now operates much like a traditional product organization.
Between dedicated product teams and a UX Center of Excellence that constantly assesses what's working and identifies which agents need to be built through front-line user research, BCG feeds those insights back to product owners, who aggregate that feedback into a centralized crowdsourcing pipeline.
“In short, we're doing everything you'd expect a product organization to do,” Wilder said.
