He left private equity to give the spreadsheet crowd a superpower.
At VectorShift, the hard part of artificial intelligence isn't the model. It's everything wrapped around it - the security, the data, the plumbing. Leonardi turned that mess into a canvas you drag and drop.
Ask most people what's hard about building with AI and they point at the model - the math, the magic, the thing that writes the sentence. Alex Leonardi points somewhere less glamorous. The connectors. The permissions. The live-syncing knowledge base that has to stay fresh while a hundred employees hammer it. The audit trail a compliance officer will actually accept.
That unglamorous layer is VectorShift, the company Leonardi co-founded in 2023 and runs as CEO. It is a no-code platform where a team can build an enterprise-grade AI application by dragging components around a canvas - a search engine, a chatbot, a document generator, an automated workflow - and ship it without writing a line of code. For the people who do write code, there's a Python SDK in the same box. The analyst and the engineer use the same product. That is not an accident.
"There are other tools, but our main differentiation is the flexibility and readiness for enterprise use cases."
Leonardi came to this problem from the inside. At Blackstone he was a private-equity data science analyst - the person evaluating transactions, deploying machine learning into portfolio companies, and helping one of the largest investors on earth figure out what generative AI could actually do. He saw the gap up close: enormous appetite for AI, almost no easy path from idea to something running in production that legal would sign off on.
So he built the path. VectorShift went through Y Combinator's Summer 2023 batch and raised a $3 million seed round in February 2024 from a roster of enterprise-minded backers - 1984 Ventures, Defy.vc, Formus Capital, 468 Capital and Y Combinator itself. The pitch was direct: give all organizations the ability to build enterprise-grade AI applications, not just the ones with a machine-learning team on payroll.
His co-founder, Albert Mao, came at the same problem from the other entrance - McKinsey, where he advised Fortune 500 companies on digital transformation and enterprise software strategy. Two Harvard grads, one fluent in the model side, one fluent in the boardroom side, both staring at the same expensive bottleneck. The division of labor writes itself.
What's striking about Leonardi is how early he got reps at running things. Before the startup, before Blackstone, he ran the Harvard College Consulting Group as CEO and grew it more than 200% year over year, past $1 million in annual revenue. It was a student club on paper. He ran it like a company. The instinct to take something small and make it operationally serious clearly predates the cap table.
And the academics are almost comic in their tidiness: a joint Harvard degree in statistics and computer science, graduated with Highest Honors at a 4.0, followed by a master's in computational science and engineering, also at a 4.0. Two degrees, no blemishes. The man does not appear to leave points on the board.
VectorShift has since sharpened its aim. The general no-code automation platform still exists, but the company has pointed itself squarely at the world Leonardi knows best: private markets. The newer framing is an AI operating system for private market investors - capturing the institutional knowledge buried in deals, diligence findings and decisions, then putting it to work on data room analysis, investment committee memos, portfolio monitoring and LP reporting. It is, in a sense, Leonardi building the tool he wishes he'd had at Blackstone.
That is the through-line. He spent his early career being the human who translated AI for an institution. VectorShift is the bet that the translator can be software, and that the software can belong to everyone - the tax advisor, the operations lead, the deal team - not just the firms that can afford to hire someone like him.
Build search engines, chatbots, document generators and workflows by arranging components on a canvas. The barrier to enterprise AI drops from "hire a team" to "open the editor."
Live-synced knowledge bases, advanced retrieval, and the security posture enterprises demand. The differentiation isn't the AI - it's being ready for the messy reality of real companies.
An AI layer for private market investors: capturing institutional knowledge and applying it to data room analysis, IC memos, portfolio monitoring and LP reporting. The tool Leonardi wished he had at Blackstone.
"Give all organizations the ability to build enterprise-grade AI applications."
- The VectorShift mission, in one line
"There are other tools, but our main differentiation is the flexibility and readiness for enterprise use cases."
Two Harvard degrees, both earned at a 4.0. The undergrad came with Highest Honors. He does not appear to leave points on the board.
He ran a Harvard student consulting group like a real business - 200%+ growth a year, past $1M in revenue - long before he had a startup of his own.
He came from Blackstone, his co-founder from McKinsey. Same enterprise-AI bottleneck, approached from finance and consulting at once.
VectorShift ships a drag-and-drop canvas and a Python SDK in the same product, so the non-coder and the developer never have to pick different tools.
At Blackstone he helped one of the world's largest investors figure out generative AI - then left to build the tooling for everyone else.
VectorShift narrowed toward private markets - the exact world Leonardi knows best. The product is, in effect, his old job turned into software.