Artificial intelligence tools such as ChatGPT or Claude are everywhere, accessible to businesses of all sizes. In Newton, a growing number of entrepreneurs, consultants, and organizations are discovering that access is the easy part. The harder question is how to leverage AI in ways that actually change how people work, without losing their sense of identity and agency.
Longtime Newton residents Kevin Gulley and Brendan McSheffrey started The Kendall Project as one way to answer that question. Their startup works with businesses trying to navigate the rush to adopt AI. It helps organizations apply the technology in practical ways while staying grounded in their internal culture and day-to-day operations. Gulley told Fig City News that while large language models know a great deal about the world, they know nothing about how a specific company actually operates.
“You have to explain how you do things,” Gulley said. “You have to deliver context.”
As an example, Gulley described working with a company whose workflow revolved around a shared spreadsheet. To employees, the document was straightforward and its use in their workflow was second nature. However, while an AI system might understand a document it might not understand how it may be used in a group.
In workshops led by the startup, employees are asked to walk through their work – step by step – and to name the tools, data, and people they rely on, along with the problems that slow them down.
By gathering insight across roles, the team can start to see which problems matter most. Employees may clash over a company’s long-term direction, he said, but they are far more likely to agree on the everyday obstacles they encounter.
“Everybody feels like they’ve been heard,” Gulley said.
The group also runs public workshops tailored to specific roles, including chief marketing officers and chief data officers.
‘A town of brainiacs’
Paul Baier said AI is flooding into the workplace whether organizations are ready or not. His focus, he said, is ensuring that Massachusetts workers and the region’s economy can evolve with it.
A Newton resident, Baier is the founder of AI Blueprint for Massachusetts, a volunteer-led initiative directed toward strengthening the Commonwealth’s AI ecosystem.
Baier said the Greater Boston area has no shortage of technical talent, driven by its universities, research institutions, and steady flow of graduates. The problem, he said, is that too many people working on AI-related challenges remain siloed.
“We’re a town of brainiacs that aren’t connected enough,” Baier said.
Baier distinguished between AI creation and AI implementation. He said many people assume AI talent refers to highly paid engineers developing large language models, but that view misses the broader range of skills companies need right now. He said companies need people who can apply AI across a range of functions, from product management and sales to customer service and other operational roles.
“This intermingling of how subject matter experts or department level experts use AI on teams is the scarce talent,” he said.
Baier emphasized how quickly the landscape is shifting, noting that what organizations are using today will look very different in six months.
“Companies are going to be managing human and AI-cognitive workers,” he said. “It’s that profound.”
Rapid changes
Angela Pitter is the founder and CEO of LiveWire Collaborative, where she guides mid-sized businesses and nonprofits through digital transformation with AI-powered marketing strategies, LinkedIn programs, and virtual events. A Boston University-trained engineer with 25 years leading global tech teams, she now helps companies integrate AI tools while establishing clear governance principles.
Pitter said that while many leaders recognize the potential of generative AI, the speed of its development has made it difficult for organizations to understand how to adopt the technology responsibly and effectively. She said she witnessed similar challenges when companies were first learning how to use social media, but now it is AI that is much more rapid in its pace of change.
“It’s almost like we’re building the airplane while we’re flying it,” Pitter said. “Nobody knows the end game here, because it keeps changing.”
Pitter said that while discussions about AI are often aligned with keeping a “human in the loop,” she prefers to think of it as keeping the “human in the lead.”
She said AI tends to be framed in dystopian terms – as a technology that will replace people entirely – when in reality it still depends on human judgment. AI tools, she said, can get organizations 95% of the way there, but reaching the final 5% of quality requires human expertise and leadership.
“We have to have a game plan,” she said.
Niche products
Kara Peterson and her husband Richard DiBona co-founded Descrybe, a niche AI product designed to assist legal professionals.
Peterson said the main thing that differentiates her company from general-purpose large language models is that Descrybe’s models are trained on primary U.S. law datasets, which give legal professionals the most trustworthy information.
“The fact that general-purpose models can access so much information is exactly the problem,” Peterson said. “That includes a lot of garbage.”
Legal work involves constant reading, analysis, and interpretation, Peterson said, and Descrybe is designed to streamline pattern recognition so attorneys can focus on higher-level thinking and client work.
“It frees up the lawyer or the attorney to do more of the higher level thinking, client work, and potentially take on more clients,” she said.
Peterson said Descrybe was initially intended to help everyday users navigate legal questions, but the company has now shifted to serving legal professionals. She cautioned against people using the platform to represent themselves.
Running Descrybe, Peterson said, has underscored how quickly AI companies have to adjust as customer needs and use cases evolve.
Chatbots
Elisa Friedman also has experience working on the enterprise side of AI.
Friedman, a Newton resident and longtime content and learning design leader, helped build Harvard Business Publishing’s chatbot Ask AI, designed to help readers of the Harvard Business Review explore leadership and management content through conversational AI on the Harvard Business Review website. Built in 2023 and released the following year, the tool relied on a curated repository of Harvard Business Publishing material rather than information pulled from the open web.
“It is a bit of black box, so you need to be very careful when you’re using it, especially if you’re creating a feature that is going to be used by hundreds of thousands of people,” Friedman said.
Friedman said building the tool required careful attention to voice, tone, and governance to safeguard the trust of the Harvard Business Publishing brand.
“We had to make sure to safeguard the trust of the brand,” she said. “We had to make sure that the language and everything was precise and we weren’t making recommendations that were just based on what I call the Claude brain.”
Governance
Chitra Sundaram, a Massachusetts-based data analytics executive at Cleartelligence, works with companies navigating the gap between AI ambition and realistic solutions.
In her experience, many AI initiatives fall short not because the technology is flawed, but because the data systems beneath them are not ready. Companies often rush toward large language models (LLMs), she said, without first addressing data infrastructure and quality.
“If you don’t govern what is being fed into the pipeline to be fed into LLMs…what comes out is garbage,” she said.
Sundaram said the rush to act often leads organizations to overengineer their approach. AI gets layered onto problems that could be solved more simply, with complex models deployed to answer straightforward questions. The result is systems that consume time, money, and energy without adding much value.
She also pointed to a shift already underway, as AI moves beyond tools that simply generate responses toward systems that can carry context, remember prior steps, and act with a degree of autonomy.
Newton in a new world
Paul Baier argued that Newton should develop a local AI blueprint – a shared vision for how the technology fits into the city’s future.
“People come to Newton for schools, the location, and the quality of life,” he said. Becoming more AI-forward, he added, could help the city grow with the times while building on its existing strengths.
Baier said Newton’s proximity to Boston’s research and innovation hubs, combined with its strong civic institutions, positions it well to play a larger role in the region’s AI ecosystem, particularly if local leaders lean in. He pointed to opportunities for the city to thoughtfully incorporate AI into its infrastructure, from internal operations to public-facing services to education.
Baier compared this moment to sitting on a beach under clear blue skies, with a hurricane forming on the horizon. Some people, he said, assume it will pass.
“It’s here now,” Baier said. “The only question is, how do you adapt?”
