Author: Lincoln Wang | Founder of MindsLeap | Global Partner at Founders Space | Founder of Founders AI Club
If you want to know whether a chef can make a great plate of shredded potatoes, the most reliable method may not be reading a resume. It may be giving the chef a simulated kitchen.
Griting AI founder Josie used this analogy in our full Founders Talk conversation. It points to a larger shift: in the AI era, companies need to move from static descriptions of past experience toward evidence of how people work in realistic environments.
The full interview transcript moves from Josie's decision to take leave from her Dartmouth AI PhD to build a company, through Talent Trust, AI Talent Audit, a consumer Learning System, a “different LinkedIn,” FDE, OPC, hiring judgment, and the operating reality of a five-person AI startup.
This was not only a conversation about one product. It was a conversation about how companies should decide whether they can trust a person's ability to create results.
From Liberal Arts to AI, Why She Chose to Build
Josie is the founder of Griting AI and is pursuing an AI PhD at Dartmouth. She had been working, studying, and building a company at the same time. As delivery became heavier and the company continued moving from zero to one, she decided to take leave from the program and focus on the next mountain in front of the company.
Her own transition from liberal arts to computer science and AI led her to two questions: what kind of talent do enterprises trust, and how can a person find the right company and career path?
Griting AI began as an attempt to turn those questions into a system. Josie did not want to build another hourly consulting or career-coaching service. She wanted to use products and Agents to help more people understand their capabilities, direction, and fit with real work.
When Agents Enter the Workflow, Talent Trust Changes
One word kept returning throughout the conversation: Agentic.
Josie sees Agents becoming a central part of enterprise intelligence. Agents are execution systems, and people will work alongside them. Companies therefore need to ask a new question: who can judge, verify, and take responsibility for work produced with Agents?
Josie describes Griting AI as a product-driven company using quantitative and engineering approaches to address a problem that has historically relied on intuition and emotion.
Talent Trust is moving from “I believe what you say you can do” toward “I can see how you perform in a real task.”
AI Talent Audit Turns an Interview into an Evidence Chain
Griting AI's flagship product is the AI Talent Audit System. Audit here does not mean financial auditing. It means taking a person's ability seriously and evaluating it through traceable evidence.
Candidates leave dynamic interaction data in the system. The system looks not only at the final outcome, but also at judgment, iteration, and behavior along the way.
The simulated-kitchen analogy makes the logic concrete. If you want to find a chef who can make excellent shredded potatoes, let ten or one hundred people work in a comparable environment. Observe preparation, execution, adjustment, and the result.
For AI engineers, generating code is only the beginning. Enterprises also need to know whether the work is trustworthy, production-ready, safe, scalable, and repeatable.
Josie described a case where a candidate's resume and GitHub looked impressive and the initial evidence seemed complete. Further checking revealed a volume and pattern of GitHub activity that did not match the candidate's background. Once the evidence chain contradicts itself, the system has to validate further instead of treating a polished signal as capability.
This does not mean AI replaces people. Experienced interviewers can notice a glance or a micro-expression. Systems can reduce the influence of emotion and personal preference. The stronger model combines an objective third-party assessment with human judgment.
From Hiring Assessment to a Learning System
Griting AI also extends the assessment logic into learning and talent incubation.
Josie put it simply: if we know how to assess, we should know how to learn.
Once an assessment can identify capability gaps, thinking patterns, and delivery problems, it can generate a more useful learning path. Josie initially imagined serving North American universities and companies undergoing AI transformation. In practice, universities have their own enrollment and employment challenges, larger enterprises often build internal systems, and smaller companies may not have enough budget.
The team therefore began shaping a consumer-facing adaptive learning and talent incubation system, while controlling cost and traffic across different product layers. Josie does not want people to pay because anxiety and fear have been manufactured around them. She wants users to be certain that they want to learn and are willing to invest in the process.
The system can also connect back to enterprises. When a learner demonstrates strong potential, Griting AI can recommend that person to a company. Hiring, learning, and talent incubation become connected rather than isolated processes.
A Different LinkedIn and the Value of a Neutral Third Party
When describing Griting AI's long-term product vision, Josie used a vivid phrase: a “different LinkedIn.”
LinkedIn relies heavily on self-reported information, connections, and endorsements. Griting AI wants to create a third-party trust system closer to demonstrated capability.
The company sits between talent and enterprises, with both sides paying, so neutrality matters. If the system identifies high-potential talent, it can recommend that person directly to a company rather than manipulating an information gap between the two sides.
Josie said the current selection process is roughly one person chosen out of three. The business model resembles a recruiting fee, but the company also takes responsibility for the outcome with a short replacement or support period when the fit is not right.
That is a broader shift in the AI era: enterprises are willing to pay not only for a service process, but also for verifiable results. A small team can compete with larger organizations through speed, flexibility, and judgment.
FDE Means Consulting Must Start Owning Outcomes
The conversation later turned to FDE, or Forward Deployed Engineer.
Josie sees FDE as a major change in the AI era. A real FDE does not simply hand tools to a customer. The role enters the customer's operating context, coordinates several Agents, solves a concrete problem, and delivers an outcome.
She offered a direct example. A company might once have hired separate UI, UX, React, and WordPress engineers. An FDE who can coordinate several Agents may be able to cover a much broader piece of work. The change is not simply fewer roles. It is a wider range of work that one person can own.
FDE also raises the bar for people. Someone may be technically strong but still unable to deliver if they lack curiosity, business understanding, and the ability to see how a company creates value.
This changes consulting. Consulting firms will not disappear, but they may need to move from knowledge and reports toward deeper domain focus, longer-term trust, and responsibility for results. Josie described the less glamorous reality of entrepreneurship: staying late to fix a presentation, absorbing a sudden client problem, and doing whatever is required to deliver.
Trust is not created by a previous title. It is created by being willing to use limited resources to make the customer's result happen.
From 0 to 0.7, an AI-Native Team at Work
Josie describes Griting AI as moving from “0 to 0.7.” The team already has real delivery, orders, and repeat business. It is now moving from product validation toward distribution and scale.
There are five people on the team, most with engineering backgrounds. Everyone is expected to master AI. The team uses tools such as Codex and Claude Code for development, alongside products such as Notion, Figma, and Canva. The important point is not a single tool. It is making AI part of how every role works.
Most of the company's current customers are in the United States, and the team is also connecting with model and cloud ecosystems through Partner Programs, training, services, and compute support. Asian expansion is a longer-term possibility, but it would require adapting to local ecosystems, product expectations, and hiring regulations.
Josie also discussed AI anxiety. AI can displace some work, but it can also create new infrastructure, energy, and real-world roles. People often see the jobs that disappear more easily than the jobs that have not yet been imagined.
The underlying enterprise hiring logic has not changed: a company needs someone who can do the work and create value. What has changed is that AI-generated resumes, interview assistance, and automated portfolios can distort the signal. Companies need stronger methods for seeing real capability.
OPC, Self-Awareness, and the Long View
Josie is cautious about the One Person Company idea. Small teams can become dramatically more efficient, but great companies are rarely created by one person alone. Products need more perspectives, and enduring companies require shared conviction. Even Steve Jobs did not build Apple alone.
When hiring, Josie ultimately emphasizes clear self-awareness and responsibility: knowing what you can and cannot do, understanding the company's situation, and being able to refuse a task that you cannot carry. Knowing your boundary is the beginning of reliable collaboration.
She still values both returning to the PhD and building the company into something larger. Entrepreneurship is a life dream, but the PhD is also a commitment. There is no simple answer here, only a continuing choice among a team, customers, contracts, and long-term promises.
Her closing advice to founders was direct: do the work well, adjust quickly and continuously, do not be afraid of change, and do not try to deceive people. As AI-generated content becomes more capable, reliable behavior and long-term trust become more valuable, not less.
Final Thoughts
The full conversation revealed more than an AI talent assessment product. It revealed an emerging framework for organizational judgment.
Enterprises will need to see evidence of real capability, understand how people work with Agents, and take responsibility for outcomes. Hiring, learning, consulting, and organizational transformation may reconnect along one operating chain.
The question is not only whether a company can find more AI talent. It is whether the company can see how a person works, how that person collaborates with Agents, and whether the person can create sustained results in the business.
That is the central judgment I took from this full conversation with Josie: in the AI era, talent does not only need to be discovered. It needs to be understood in a real environment.
About Founders Talk
Founders Talk is an interview program created by Lincoln, founder of MindsLeap. It invites influential founders, entrepreneurs, AI experts, investors, creators, and global innovation practitioners in the AI era to discuss frontier technology, business models, enterprise AI transformation, and cross-market opportunities.
The program helps audiences understand global frontier technology trends and gain first-hand insight into enterprise AI transformation. It focuses on lived experience and key judgments, showing how new technologies enter products, organizations, global brands, and individual creativity.
About MindsLeap
MindsLeap is an AI-native organizational transformation acceleration platform.
Through deep collaboration with Founders Space, the Silicon Valley innovation incubator, MindsLeap connects global AI frontier thinking, the Silicon Valley technology startup ecosystem, and real transformation scenarios for Chinese entrepreneurs. Around entrepreneurs, founders, AI engineers, industry experts, and investors, MindsLeap builds Founders AI Club, AI workshops, innovation study tours, executive exchanges, and enterprise transformation programs to help organizations develop the mindset, methods, and operating capability required for the AI era.
MindsLeap focuses on how AI enters organizational capability, business workflows, product innovation, and growth systems. Starting from the first practical use case, we help companies develop reusable workflows, digital employees, and AI-native teams so that transformation can move into real business and produce measurable results.
This article was translated and adapted from the Chinese original with AI assistance.