The Profile · Enterprise AI · Vol. VII
Saurabh
Jain
The founder who decided that skills are a graph, not a spreadsheet, and then spent fifteen years building the graph.
The founder who decided that skills are a graph, not a spreadsheet, and then spent fifteen years building the graph.
Saurabh Jain runs a company that thinks about your job the way a cartographer thinks about a coastline: as something with edges, adjacencies, drift, and a persistent tendency to be mis-drawn. Spire.AI, the company he founded in 2007 and still runs, sells enterprises a Large Graph Model for skills - a phrase that sounds like something a consultant made up until you notice that most of the alternatives are, literally, spreadsheets.
Spire.AI's pitch is that most HR software has been asking the wrong question. The wrong question is: what does this employee have on their resume. The right question, according to Jain, is: what can they actually do, what are they adjacent to, and how quickly could they get to the thing you actually need them to do next quarter. The first question fits in a spreadsheet. The second one is a graph.
The company describes its core primitive as a Large Graph Model, or LGM - a play on LLM that is either clever or annoying depending on how many acronyms your day has already contained. The point is that skills are not a flat list. They cluster, they decay, they have neighbors. "Python" is closer to "R" than to "conflict resolution." A senior data engineer with three years of Snowflake is one weekend of study away from being useful in a Databricks shop; that fact does not appear anywhere in her HRIS record, but it exists in the graph. Spire's argument is that if you can model that graph well, you can do a lot of things that HR software has historically failed at: match candidates to roles, spot internal mobility opportunities, forecast reskilling needs, and, if you are ambitious, make the workforce planning conversation resemble something a CFO would recognize.
That last point is one Jain returns to often. In a 2024 essay in The People Management, he argued that the CHRO and the CFO should be treated as equal partners - not a common talking point from a founder whose customers are, mostly, CHROs. It is the sort of position that plays well in boardrooms and less well in departmental turf wars, which is probably why he keeps saying it.
The newer wrapper is OpenKnowra, which Spire positions as an enterprise context intelligence layer sitting under the talent copilot. The distinction, roughly: Spire.AI knows what your people can do; OpenKnowra knows what your enterprise is. If Spire's original bet was that skills were a graph, OpenKnowra's bet is that everything else - operations, customer state, product knowledge - is a graph too. Whether the market wants two graphs from one vendor is one of those questions that gets answered in about eighteen months.
The company sits inside Spire Innovations Inc., headquartered at a Woodbridge Township, New Jersey address that reads like a lot of enterprise software addresses read: unremarkable in a way that suggests focus. Jain himself is based in New York. About 140 people work at the company. Total disclosed funding is around nine million dollars, a Series A of roughly eight million logged in 2013, which is either a rounding error by 2026 AI-funding standards or a sign that the company has been operating without needing much outside money for a long time. Both readings are probably true.
The phrase Spire uses most is "domain-intelligent." It is doing real work. The claim is that a generic large model, dropped into HR, will confidently confuse a "senior software engineer" at a bank with one at a game studio, because it does not know the domain. A domain-intelligent model, in Jain's telling, is trained and structured to notice that these two engineers are in different graphs entirely, with different adjacent skills, different career pathways, and different unit economics. The category is fuzzy enough that plenty of competitors also claim it. What Spire has, that most do not, is fifteen years of running the same experiment on real enterprise customers.
Which brings up the second unusual thing about Jain: he has been running one company since 2007. In a software category where founders tend to churn through three ideas before their kids finish middle school, this is a data point. Companies that survive fifteen years without a marquee exit are usually either quietly profitable or quietly stubborn. Spire.AI, from the outside, looks like both.
AI isn't just a buzzword in talent acquisition; it's a game-changer.- Saurabh Jain, interview with DataQuest India
The teal edges are the ones Spire.AI cares about: not what an employee already is, but what they are two weekends away from.
Jain's most-repeated non-product argument: workforce economics should sit as high in the org chart as capital economics. He wrote a whole essay on it.
A resume is a snapshot. A skill graph is a system with edges that grow, decay, and cluster. The distinction is not academic; it is what the software actually models.
Generic models confuse a bank engineer with a game engineer. A domain-intelligent model, in Jain's telling, knows they live in different graphs entirely.
Fairness, transparency, accountability - phrased less as boilerplate and more as constraints on how the copilot ranks and recommends.
Companies keep hiring outside for skills they already have inside. The graph makes those adjacencies visible; a spreadsheet cannot.
Every HR conference talks about "skills-based organizations." Jain's line is that most rollouts fail because the underlying data model is still a taxonomy. Fix the model, then the org.
Harnessing AI in talent acquisition and management is not just about adopting new tools; it's about transforming the organizational approach to skills and talent.- Saurabh Jain
There is a small, telling detail in Spire.AI's public footprint. The company's Twitter handle is @spire2grow. Its LinkedIn slug is spireai. The founder's personal LinkedIn slug is /in/spirian. This is a founder who has committed - on a level that would be considered aggressive branding at a consumer startup - to the word Spire, in a category that mostly names itself after acronyms like ATS and HCM and TXM. It is a small aesthetic choice, but it says something about the company. Spire.AI is trying to sound like an idea, not a workflow.
The other detail worth noting is educational. Jain's advanced degree is from the GDW Consortium, a joint program that runs across IIM Bangalore, UCLA Anderson, SDA Bocconi, and City University of Hong Kong. It is the sort of credential that sounds invented until you check, and then it turns out to be a genuinely unusual four-continent operator's education. Most founders who cite an MBA cite one school. Jain, in effect, cites four. The pattern-recognition here - global enterprise, cross-jurisdictional, comfortable with complexity - lines up neatly with what Spire.AI actually does, which is sell a graph model of the workforce to multinationals that have never been able to see their own workforce clearly.
The company's geographic setup rhymes with that. The corporate address is in Woodbridge Township, New Jersey, an easy commute to New York. The founder lives in New York. A large share of the engineering DNA runs back through India, where Jain's Yahoo and Apigee years happened, and where a lot of enterprise HR product engineering still gets built well. This is not a "we're a Bay Area startup" company. It is an East Coast enterprise software company with a serious India engineering backbone, in a category where that combination has historically shipped more than it has hyped.
Fifteen years in, Spire.AI is not the sort of company that shows up in TechCrunch every third week. It shows up in HR trade press, on Forbes Council mastheads, and in interviews where the interviewer asks a very specific question about talent supply chains and gets a very specific answer back. There is a category of founder whose talent is not for narrative but for staying in the same conversation for a decade and a half until the conversation eventually turns their way. Skills-based organizations, internal mobility, workforce graphs - these were niche vocabulary in 2010. They are boardroom vocabulary now. Jain has been running the company that treats them as first-class primitives the entire time.
Whether that patience gets rewarded depends, mostly, on whether the enterprise buyer of 2026 is finally ready to admit that they cannot see their own workforce clearly and that a graph might help. The evidence is mixed. What is not mixed is that Spire.AI has been ready for this moment longer than almost anyone else selling into it. Which is either an early advantage or a very long wait, depending on the quarter you ask.
Jain, for his part, keeps writing. He publishes on Spire's thought leadership pages. He publishes on Forbes. He publishes on YourStory. The essays are dense in the way founder essays usually are not - closer to a systems engineer's whitepaper than to a LinkedIn thought-leadership post, which is what makes them worth reading. Read enough of them in a row and the underlying worldview lines up: enterprises are systems, workforces are graphs, skills are edges, AI is the thing that lets you finally see the shape of what you already had.
That is, more or less, the whole pitch. It has taken him fifteen years to be able to say it in a room and have people nod instead of squint. He is, at the moment, being nodded at more.