MIT Sloan professor Rama Ramakrishnan has spent more than two decades building AI systems for companies like Salesforce, Oracle, and McKinsey. Now, in the final session of his three-part masterclass on artificial intelligence, he's focused on how AI is accelerating scientific research and drug discovery - and how researchers can use it to solve problems that traditional computing can't touch.
Ramakrishnan, who is Professor of the Practice in Artificial Intelligence and Machine Learning at the MIT Sloan School of Management, explained that AI is helping scientists move faster across multiple fields. The payoff is highest in life sciences, where AI can identify potential therapeutic compounds, improve clinical research, and reduce the time and cost of bringing new treatments to market. But he expects similar benefits across any discipline that depends on processing large datasets or identifying patterns.
AI in drug discovery
Drug development is notoriously slow and expensive. Ramakrishnan said AI is changing that by identifying promising drug candidates faster than traditional methods and by helping researchers analyze data that would take years to uncover without sophisticated algorithms.
"AI is not simply changing how we work; it is changing how we discover, learn and solve some of society's most complex challenges," Ramakrishnan said.
That shift comes with a caveat: AI works alongside human expertise, not instead of it. The value, he said, comes from pairing AI's data-processing power with a researcher's domain knowledge and judgment.
From Salesforce to the classroom
Ramakrishnan brings product-building experience to the academic conversation. Before MIT Sloan, he spent two decades founding and scaling software companies that were later acquired by Oracle, Demandware, and Salesforce. At Salesforce, he was chief data scientist of Salesforce Commerce Cloud, where he led the development of Einstein for Commerce, an AI platform that delivers personalized shopping experiences to hundreds of millions of consumers each month. Earlier, he worked at McKinsey & Company.
That background shapes how he teaches. At MIT Sloan, where he teaches predictive and generative AI, he focuses on practical, usable skills - not theory for its own sake. His materials are also on MIT OpenCourseWare, and he writes a column for MIT Sloan Management Review.
Practical guidance for researchers
Ramakrishnan's larger point: AI tools are useful for more than drug discovery. He said they're increasingly valuable for any researcher or professional who works with large volumes of information, needs to spot emerging trends, or wants to base decisions on evidence rather than guesswork.
For someone looking to integrate AI into their daily workflow, Ramakrishnan suggests a modest approach: treat AI as a collaborative partner, not as a replacement for expertise. Use it to analyze data, surface patterns, draft content, or run a first pass on a problem. Then apply your own judgment to results. That's how AI stays a tool in the lab, rather than a black box.
Why this matters for Science & Research professionals
The role of the researcher is shifting. Reading stacks of papers or analyzing datasets that take months to parse is still labor that only a human can do - but expect AI to take over the most repetitive and time-consuming parts of that work. Ramakrishnan estimates that scientists who build AI into their research workflow will be able to ask questions that were impossible to tackle before, just because the data analysis component just got an order of magnitude faster.
That doesn't mean a scientist's judgment becomes obsolete. It's the opposite: the value of a skilled researcher goes up when AI handles the brute-force analysis. For scientists thinking about the next few years, the practical question is less about whether to use AI and more about where in the pipeline it makes sense comes into their specific research process.
Your membership also unlocks:
