Kian Katanforoosh has one of the best mentors in the AI world: renowned researcher Andrew Ng, who also served as his graduate school advisor at Stanford. The two went on to create Stanford’s deep learning program, and now, Ng serves as Chairman of Katanforoosh’s startup, Workera. Ng has been a meaningful guide in Katanforoosh’s career, but now, the Workera CEO is trying to break down what it means to be a good mentor, and automate it through a new AI agent, Sage.
“I trust Andrew because I understand his background and expertise, but how many Andrews are there in the world? Not that many,” said Katanforoosh in an interview. “So automating this mentorship aspect was critical.”
On Tuesday, Workera announced Sage, an AI agent you can talk with that’s designed to assess an employee’s skill level, goals, and needs. After taking some short tests, Workera claims Sage will accurately gauge how proficient someone is at a certain skill. Then, Sage can recommend the appropriate online courses through Coursera, Workday, or other learning platform partners. Through chatting with Sage, Workera is designed to meet employees where they are, testing their skills in writing, machine learning, or math, and giving them a path to improve.
To be clear, Workera’s definition of “mentor” is a hyper-specific one; Katanforoosh acknowledges that Sage will not do everything Ng did for him. Encouragement, career guidance, and networking are hard to automate. But Sage can somewhat objectively assess an employee’s skillset, and recommend the right courses to help achieve their goals. While that’s not a perfect mentor, it’s better than some have access to.
As the son of Iranian immigrants, Katanforoosh’s parents were forced to flee their home country during the unstable 1970s revolution, abandoning their studies in the process. His father left a science degree behind and ended up selling clothes in France to make a living. While today’s Workera largely serves Fortune 500 employees, Katanforoosh believes that making assessment skills more available could one day help people in his parents’ situation.
People are craving mentorship today more than ever. In the era of remote work, young employees receive less face time with experienced colleagues, meaning less chances to catch some nuggets of wisdom by the water cooler. The CEO of Workera thinks the company’s new AI agent is up to the task.
Sage will roll out to early access customers as of November 2024, including military contractor Booz Allen. Other Workera customers, including the U.S. Air Force and Accenture, will get general access to Sage in March 2025.
Workera has raised more than $44 million to administer AI-generated tests for enterprise employees, offering employers a way to gauge their employees’ skillsets. Employees typically don’t jump at the chance to be compared to their coworkers, but Workera tries to offer a way for businesses to invest in their employees as well.
Sage does the same thing, but offers a more conversational experience that ties together the Workera platform more neatly. The flexibility of OpenAI’s multimodal models, which Workera uses, also offers a flexible interface that’s capable of scaling to drastically more tasks in various mediums. 95% of Sage’s interactions are powered by GPT-4o, according to Katanforoosh, while the AI agent uses OpenAI’s new o1 model 5% of the time to plan out more complicated tests which may require some reasoning.
Can an AI agent replace a great human mentor?
With Sage, Workera is claiming to become a mentor for workers, instead of just an administer of skill assessments. A human mentor offers emotional support, encouragement, and connections that an AI chatbot likely never could. However, in some aspects of mentorships, Katanforoosh thinks Sage can be better.
“A good mentor needs to assess properly, because unless the mentor can assess accurately, it cannot help you… That’s one that can be automated,” said Katanforoosh. “In fact, I’m pretty confident measurement systems we have today are better than most people; I would trust the Workera system much more than I trust myself at measuring someone’s machine learning skills.”
Another aspect to this is that humans are biased and highly influenced by superficial traits, so they may not always make accurate assessments of someone’s talent. AI systems are not perfect either, for the record, and contain many of the same biases as the humans that created them. After all, AI models are mostly based on human-generated data.
But the biases of AI models may have a more promising solution than that of humans. Katanforoosh teaches a course at Stanford on bias mitigation methods in AI. He firmly believes there are ways to reduce bias underlying an AI model’s data with algorithms. These can weight gender, race, or other considerations in a different way, making the outputs of AI models more equitable.
“I actually feel very confident that AI is already much less biased, but will be even less biased than humans in the coming years,” said Workera’s CEO.
By automating these tasks, Katanforoosh says managers can be freed up to manage the human aspects of mentorship that AI can’t automate. Human managers still need to encourage and guide their employees, among other things that AI mentors are still limited in doing.
One thing Sage does not do yet is teach long form content. For that, Workera relies on partners in the online learning space. However, Sage will identify skills you might be able to learn quickly, and generate a brief scenario and question to test your understanding.
I would argue Workera is stretching the word “mentor” here, and somewhat is using its own definition. That said, Sage may be a useful agent that managers can add to their tool belt to help assess and invest in their workforce.