Take action immediately: The most vital lesson is to execute without delay. AI has made everything from engineering a product to launching a digital interface incredibly accessible. There is no longer an excuse for inertia. If you have an idea, pursue it, because you learn the most through direct action.
I had the pleasure of talking with Sachin Kamdar. The first software Sachin built for work was not meant to impress venture capitalists or transform enterprise productivity. It was meant to help him understand a classroom.
He was teaching math and economics at an alternative high school in Brownsville, Brooklyn, after college, standing in front of students who were all expected to reach the same state standards despite arriving with sharply different skill levels. The problem was not abstract. It was daily, human, and stubbornly practical: how do you lead a group toward one shared goal when every person needs something different to get there?
Kamdar’s answer was to build software that could identify gaps in each student’s core competencies and help him create personalized curriculums. “I have always been a technology aficionado,” he says, but in that classroom technology was not the story. It was a tool in service of teaching. The system worked well enough that the Department of Education asked him to help scale it across multiple classrooms. Then came the part that would shape him just as much as the success did. The focus, he recalls, shifted toward “politics and budgets rather than student outcomes.” Disheartened, he stepped away.
That early collision between idealism and institutional friction runs through Kamdar’s career. He is now the founder and CEO of elvex, an enterprise AI platform built for large organizations trying to turn their workforces into AI-native builders. Before that, he co-founded Parse.ly, the content analytics company used by major publishers and acquired by Automattic in 2021. His résumé can read like a clean founder arc: teacher, entrepreneur, successful exit, second act in AI. But the more revealing through line is narrower and more consistent. Kamdar keeps returning to the same problem in different forms: how to help large groups of people learn, adapt, and act intelligently without crushing them under systems meant to manage them.
He grew up in the Midwest, born in Kentucky and raised in Ohio, before moving to New York for college and staying there for more than two decades. Long before he was building enterprise software, he was tutoring students in high school and college. Teaching, he says, has been the “running thread” of his life. It carried him from volunteer tutoring to public education to analytics software and now into AI. Even his current business, despite the vocabulary of models, agents, governance, and infrastructure, is still partly a teaching problem: how do you make a powerful new capability usable for ordinary people inside complicated organizations?
At elvex, Kamdar is trying to solve what he sees as the false choice many companies face with AI. One approach is chaos: leadership buys Gemini, Claude, ChatGPT, and other tools, then waits to see what employees do. “Oftentimes, that approach fails dramatically,” he says. A few people become power users, while most employees use AI to draft emails or summarize documents. The opposite approach is command and control, with rigid top-down strategy and a thick governance layer. That fails too, in his view, because it creates so much friction that experimentation dies before it can teach the organization anything.
elvex is built for the space between those extremes. It lets employees create custom agents for specific workflows while giving companies centralized management, visibility, and governance. Kamdar describes the goal as turning “everyday knowledge workers into AI-native builders” in a way leadership can approve rather than fear. The company serves larger organizations, typically those with more than 500 employees, because the value compounds only when local discoveries can spread. If one sales employee finds the exact workflow to pull pipeline reports, update Salesforce, and trigger outbound campaigns, the point is not simply that one person became more efficient. The point is whether the organization can turn that isolated skill into a shared capability.
Kamdar’s view of AI is optimistic, but not frictionless. He is blunt about hallucinations and accuracy limits, but he thinks much of the public conversation confuses perceived risk with real risk. “Humans are not 100% accurate either,” he says. The answer, in his mind, is not to pretend AI is perfect or to avoid it altogether. It is to manage AI with the same operational controls one would apply to human work: oversight, accountability, context, and judgment.
He is equally skeptical of the idea that AI’s main business use will be replacing people. Some companies, he argues, are using AI as a convenient explanation for downsizing they might have pursued anyway. The more important story, he believes, is what happens when organizations discover how much more work becomes possible. “There is simply more to do,” he says. His clients are already deploying multiple agents per employee, and he expects that number to grow sharply. A company of 1,000 people, in his example, may soon be managing 10,000 agents. Those systems still need humans to direct them. Without that guidance, he says, they are “like a flock of sheep without a shepherd going nowhere.”
There is a teacher’s assumption inside that metaphor: tools do not become useful merely because they are powerful. They need direction, context, practice, correction, and people capable of understanding what the work is for. That belief also shapes Kamdar’s advice to founders. Take action immediately. Do not let AI make you intellectually lazy. Write without AI first, because writing clarifies thought. Hire exceptional people. Keep learning. Beneath the startup language is an older discipline: do the work yourself before you outsource your judgment.
Kamdar now lives in Oakland and is building in the Bay Area, but his argument about AI still leads back to the classroom. The question is not whether a new system can produce more output. The question is whether people and organizations can learn fast enough to use that output well. In this interview, Kamdar discusses the future of enterprise AI, why he believes adoption requires both freedom and governance, and why the workers who lean into this shift may gain more leverage, not less, in the economy being built around them.
Yitzi: Sachin, it’s a joy to meet you. Before we dive in and talk about your amazing work with AI, our readers would love to learn about Sachin Kandar’s personal origin story. Can you share with us a story of your childhood, how you grew up, and particularly the seeds of all the amazing creativity that has come since then?
Sachin: I’ll make it somewhat short. I grew up in the Midwest — born in Kentucky, raised in Ohio — and then moved to New York for college. I remained in New York for about 22 years through college and post-college, and now I live in the Bay Area out in Oakland.
There is a running thread of what has interested me over the course of my 40-plus years on this planet, and it really has to do with teaching and education. That passion has carried me through in a myriad of different ways. The straight line through that journey began in high school and college, where I volunteered to tutor students who needed extra help.
After college, I became a public school teacher in Brooklyn at an alternative high school. It was my first experience leading and guiding a group of people toward a common goal, which, in that case, was graduation and performing well on state-run tests. One of the challenges I managed in that classroom environment was teaching a group of students who arrived with a highly diverse set of skills but needed to learn the same curriculum.
I have always been a technology aficionado, trying to leverage it to assist me in various ways. In that classroom, I built software to help me identify the gaps in the core competencies of each student so I could build personalized curriculums for them. That approach performed remarkably well. I was eventually tapped by the Department of Education to help scale and run that system across multiple classrooms. However, that turned out to be a frustrating experience because the focus shifted toward politics and budgets rather than student outcomes.
Disheartened, I stepped away and reconnected with an old college roommate. Back in school, we had always talked about building companies together, and we decided to finally pull the trigger. We launched a company called Parsley Analytics, which grew into a highly successful enterprise. I ran it for the better part of a decade before selling it in 2021. Soon after, I caught the entrepreneurial itch again and began building in the enterprise space once more, choosing AI as our core underpinning. That is the nexus of how elvex got started.
Yitzi: Please tell us more about your current business. Tell us the pain point it’s trying to solve and what makes it different from all the other amazing people doing AI.
Sachin: Of course. Every company right now is dealing with onboarding and transforming their workforce with AI. This presents a massive challenge inside the enterprise space because organizations have to manage this shift across hundreds, thousands, or even tens of thousands of individuals.
What you frequently see happening inside these organizations is a tendency to throw spaghetti at the wall. From a software perspective, leadership might decide to acquire Gemini, Claude, and ChatGPT simultaneously just to see what people do with them. Oftentimes, that approach fails dramatically. A couple of individuals figure it out, while the rest of the organization uses it strictly to draft basic emails or summarize content, never truly maximizing the capacity of the technology.
The alternative approach is to set an incredibly rigid top-down strategic aim for AI adoption, controlling every variable with a heavy governance layer. That also fails because it introduces so much friction that employees cannot experiment or innovate.
The problem we solve is enabling both methodologies to succeed simultaneously. We allow people to build to their heart’s content while providing the necessary centralization, management, and governance on top. This equilibrium gives leadership the visibility and control required to let the technology truly transform the business.
That is where elvex comes in. We operate across any model you want to work with and integrate with any type of software you leverage. Most importantly, we turn everyday knowledge workers into AI-native builders in a way that is fully sanctioned and welcomed by corporate leadership.
Yitzi: Just so I’m clear, it sounds like you are leveraging existing off-the-shelf LLM models and helping people create custom software. Am I understanding that correctly?
Sachin: Custom agents. We are helping them create agents to handle specific workflows. Most of these are internally focused workflows for individual productivity. We also help teams collaborate, meaning that when you create an agent, it scales to serve one-to-many instead of remaining strictly one-to-one. Additionally, we assist them in building organizational context that compounds over time, turning every employee into a power user with localized expertise rather than leaving those insights locked with a single individual.
Yitzi: Is this an analog to something like Anti-Gravity or Codex, or is there a specific problem it is trying to solve more than those platforms?
Sachin: The closest analog would be an independent platform that works universally with any type of model. It is designed precisely for that middle ground: providing enough governance for corporate visibility while focusing heavily on an intuitive user experience so it remains frictionless for anyone to pick up and build with. That is the precise line we walk with elvex, and it is why we attract our current client base.
Yitzi: So you are saying your platform works with a wider variety of AI models compared to single-provider solutions?
Sachin: Yes. We support open-weight models alongside the latest up-and-coming systems. While platforms like Anti-Gravity are geared predominantly toward technically minded individuals, elvex is purpose-built for the standard business user and knowledge worker.
Yitzi: I love the phrase you used, AI transformation, which brings to mind digital transformation. The image that comes to mind is that eventually everyone will transform their systems from analog to digital, and then to AI. Is that where you see the world heading?
Sachin: Yes. The most successful companies are going to view AI not as an incremental tool, but as core infrastructure that the business runs on. Organizations that achieve this integration first will hold a massive competitive advantage over those that lag behind. This shift represents a true transformation. Instead of merely adding another tool to your toolkit, it involves re-architecting companies and teams to build directly on top of this foundational layer.
Yitzi: What kind of customers do you work with? Do you work with smaller businesses like a medical or dental office, or is it geared toward larger enterprise companies?
Sachin: Because of the specific architecture we built into the platform, our smallest clients typically have over 500 employees, and we scale up to organizations with tens of thousands of workers. The primary advantage of a platform like ours stems from the compounding value of turning an entire workforce into power users. If you run a small office of ten people, our system is going to be overkill because you can achieve alignment simply by talking around a conference table.
However, if you are a larger company, you want to identify the person on the sales team who figured out the exact right workflow to pull pipeline reports, update Salesforce, and initiate outbound campaigns automatically. You want to extract that capability and scale it across the entire department instead of letting it remain isolated within one person’s experience. elvex takes what is working exceptionally well locally and multiplies its value through structured team collaboration.
Yitzi: The idea of AI transformation sounds very exciting, but we are still in its infancy. Even as great as these models are, they still hallucinate. What are some of the potential roadblocks you see in AI transformation generally, and how is your company solving them?
Sachin: The roadblocks frequently stem from a gap between perceived and real risk. There is a heavy perceived risk regarding AI hallucination and accuracy constraints. The technology certainly hallucinates at times and is not 100% accurate, but market expectations are often decoupled from reality. Humans are not 100% accurate either; we make mistakes daily. Managing AI is less about avoiding perceived risk and more about establishing the same operational controls you would implement for any human employee on your team.
Concurrently, there is a broader socioeconomic shift happening. In my opinion, many companies are using AI as a convenient scapegoat for downsizing. If their stock is underperforming or they miss quarterly targets, they execute layoffs and attribute the reduction to AI efficiencies.
I firmly believe that companies recognizing the true value of AI will actually increase hiring because they need human capital to capture the upside scenarios that the technology unlocks. Ramp, the expense management company, issued a report yesterday demonstrating exactly that: their lead economist showed that the most AI-native companies are currently hiring at the fastest rate.
The faster individual employees transform their workflows to demonstrate the upside of AI, the more leverage they will secure within their organizations, standing in direct contrast to top-down, replacement-minded narratives. By adapting quickly, individuals gain the agency to show that AI creates opportunities and requires human collaboration. That is what is playing out in real life with the companies leveraging AI the most.
Yitzi: That is a fascinating point. To ensure I understand you correctly, you are saying those who leverage AI effectively are not necessarily going to downsize, but will actually hire more because they become far more prolific and discover more opportunities?
Sachin: Exactly. There is simply more to do. More opportunities appear, these businesses scale faster, and they require personnel to manage that momentum. We are observing exponential growth in the number of AI agents deployed by our clients. Right now, there are roughly 3.5 agents for every human employee, and we expect that ratio to reach tenfold within the next year.
Think about the implications: a company of 1,000 people will manage 10,000 agents. These agents cannot self-manage; they require human oversight, guidance, and strategic orientation. We are nowhere near Artificial General Intelligence (AGI) where systems operate entirely on their own, and I do not anticipate we will be anytime soon.
Consequently, forward-thinking companies require more talent. These agents expand operational capacity and run 24/7 instead of just replacing human tasks. You need people to direct that continuous output, otherwise, it is like a flock of sheep without a shepherd going nowhere.
Yitzi: That is amazing. So the key is to upskill your employees and train them to manage these agents, rather than letting them go.
Sachin: Yes. This responsibility falls on both organizations and individuals. The more an individual employee can demonstrate their ability to navigate AI, the more leverage they command in the professional relationship because they are proving the upside of the technology.
AI transformation is essential for keeping businesses competitive, and it is economically crucial for the country to accelerate this process. This shift will unlock the next wave of macroeconomic growth while ensuring the workforce is pulled forward with it. We are not moving toward a future where AI runs everything in a vacuum without employees. The empirical data clearly shows that the more you leverage AI, the more you need human expertise to direct it.
Yitzi: So there will be a net gain in productivity; even though you need more people, the overall production will be vastly higher.
Sachin: That has been the case with almost every major technological shift in human history. When an efficiency gain lowers the cost of a service or commodity, market demand spikes. As efficiency rises and prices drop, demand increases, requiring more human infrastructure to manage the scale.
The same economic principle applies to AI. The transition will not be entirely seamless; there will be displacement, role evolution, and a profound need for retraining. However, the faster we navigate that friction, the sooner the workforce will experience positive economic momentum driven by this technological change.
Yitzi: That makes sense. You are saying the best way for individual employees to ensure job security is to embrace AI, bring new capabilities to leadership, and demonstrate how to manage these new AI-driven workflows effectively. Am I understanding you correctly?
Sachin: Yes, and leadership needs to do it too. This dynamic is identical to the introduction of computers or electricity to industrial workflows. The individuals and enterprises that masterfully demonstrated the advantages of those technologies leapfrogged the competition. It revealed a broader horizon of what was possible and illuminated exactly why scaling headcount was necessary to capture the new upside. AI follows that exact trajectory. The only variable is the eventual emergence of AGI, but even then, it will likely manifest as a gradual gray area rather than an overnight shift, meaning the transformation narrative remains paramount.
Yitzi: Do you see AI as a utility, like electricity or gas, rather than just a piece of software?
Sachin: I absolutely do. It will operate at that foundational scale. We are already seeing SaaS business models evolve to reflect this reality, migrating away from rigid seat-based pricing toward usage-based metrics. Companies want to evaluate exact consumption and clear ROI instead of paying a flat fee for a user license.
Imagine if utility companies charged for electricity on a seat-based model where you paid a flat twenty dollars per employee regardless of power consumption; it wouldn’t make sense. The software industry is heading toward that exact same utility consumption model.
Yitzi: Do you imagine that in the future, nearly every SaaS platform or electronic device will have some form of AI behind it?
Sachin: Anything oriented around knowledge work or services certainly will. I am not entirely certain we need AI integrated into every physical tool, though automated lawnmowers are already emerging, so the boundaries are fluid. However, I do not see AI touching 100% of human activity. There will be distinct areas where it becomes clear overkill, much like smart refrigerators or overly complex printers that offer features consumers do not genuinely require.
Yitzi: Turning a bit afield, are there any hidden AI tools or gems that you love that aren’t getting enough attention, which you would like to share with our readers?
Sachin: First and foremost, elvex. I use our own platform for multiple hours every day, and it is an absolute joy to build a product that you personally rely on constantly.
Outside of our platform, I use a calendar application called VimCal. They existed prior to the current AI wave, but they have smartly layered in automated scheduling features that I find highly efficient. Additionally, if you have not explored real-time voice modes in AI yet, it is a highly conversational and transformative experience. I am eagerly awaiting a similarly high-fidelity upgrade to standard virtual assistants like Siri in the near future.
Yitzi: This brings us to our signature question, Sachin. You have achieved a great deal of success and learned a lot along the way. Looking back to when you first started, what are five lessons you have learned about leading an AI company, perhaps some learned the hard way, that someone starting out today could benefit from?
Sachin: These insights span both my traditional tech roots and my current AI experience:
- Take action immediately: The most vital lesson is to execute without delay. AI has made everything from engineering a product to launching a digital interface incredibly accessible. There is no longer an excuse for inertia. If you have an idea, pursue it, because you learn the most through direct action.
- Do not let AI induce complacency: The technology can easily lead to intellectual laziness if you stop applying critical, first-principles thinking. AI can assist you with optimization, but it cannot replace the core analytical work required to navigate complex business challenges.
- Write without AI first: Writing remains one of the single best exercises for clarifying your thoughts. Try drafting your strategic ideas independently first, then leverage AI tools to polish the phrasing afterward.
- Prioritize exceptional talent: Even in an AI-driven ecosystem, businesses are built by people. Be highly selective with your hiring practices. Finding exceptional talent ensures you have a team capable of solving any problem, regardless of technological shifts.
- Maintain a student mindset: Never stop learning. No matter how much success you achieve, you must remain observant and adaptable. This agility is more critical than ever given the unprecedented pace of technological advancement.
Yitzi: How can our readers try out elvex and experience it for themselves?
Sachin: We do not have a self-serve offering live just yet, though we plan to test one in the coming months. In the meantime, anyone interested can reach out to me directly via email at sachin@elvex.ai, and I will be happy to personally set you up with access.
Yitzi: This brings us to our final question. Sachin, given the platform you have built, you have a significant opportunity to influence the industry. If you could spread one idea or inspire a movement to bring the greatest amount of good to the most people, what would it be?
Sachin: It aligns with what we discussed earlier: the faster individuals embrace this transformation and demonstrate to employers how much they can accomplish with AI, the more economic leverage they will secure. Everyone should lean in to discover their potential and highlight these capabilities. This shift will ultimately reinforce the understanding that we need more people to drive new initiatives, rather than scaling back human potential.
Yitzi: Beautiful. How can our readers continue to follow your work, engage your services, and support your mission?
Sachin: You can follow me on LinkedIn under Sachin Khandar, where I post regularly. We also publish a newsletter through elvex that you can subscribe to, and we host a podcast called Building for Others, where we interview individuals creating practical AI tools for widespread use.
Yitzi: Sachin, thank you so much for your time. It has been an honor to connect, and I wish you continued success.
Sachin: Thank you, Yitzi. It was great to meet you. Take care.
Yitzi: You too. Goodbye.
How Sachin Kamdar Is Building the Future of AI Workforce Transformation was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.