BitBoard Is Betting That the Spreadsheet of the Future Is a Conversation

The Macro: The Spreadsheet Is 45 Years Old and Still Runs Everything

VisiCalc shipped in 1979. Lotus 1-2-3 took over in 1983. Excel won the war in the 1990s. Google Sheets added real-time collaboration in 2006. And then… nothing. The spreadsheet has not had a fundamental interface innovation in nearly two decades. We got better formulas, more rows, some charting improvements, and cloud sync. But the core interaction model is the same: you click into cells, you type values, you write formulas, you drag to fill.

This would be fine if spreadsheets were a niche tool. They are not. Spreadsheets run the world. Financial models, project plans, inventory tracking, HR databases, marketing analytics, sales pipelines. Somewhere in every company, there is a spreadsheet doing critical work that probably should have been a proper application years ago.

The problem is that spreadsheets are simultaneously the most powerful and the most frustrating tools in business. Powerful because they are infinitely flexible. Frustrating because that flexibility comes with a steep learning curve. VLOOKUP. INDEX MATCH. Pivot tables. Array formulas. Named ranges. The gap between what spreadsheets can do and what most people know how to make them do is enormous.

The AI spreadsheet space is heating up. Rows.com uses AI to automate data tasks. Equals built a spreadsheet for startups with built-in connections to business tools. Airtable has been adding AI features to its database-spreadsheet hybrid. But most of these tools bolt AI onto an existing spreadsheet paradigm. The AI helps you write formulas or clean data, but the fundamental interaction is still “click cells, type things.”

What if the spreadsheet itself was built around a conversation?

The Micro: From Palantir and BlackRock to a Blank Sheet

BitBoard was founded by Connor Jones and Ambar Choudhury. Connor is an ex-Forward, BlackRock, and Columbia Engineering alumnus. Ambar was the first engineer at Forward, the AI-enabled healthcare company, and previously worked at Palantir. Both of them have spent their careers building tools that help people work with complex data. They are part of Y Combinator’s Spring 2025 batch with David Lieb as their group partner.

The product is an AI-first spreadsheet where the primary interaction is chat. You describe what you want to build, analyze, or change, and an AI agent handles it. Want a financial model? Describe the assumptions and the agent builds it. Want to analyze a dataset? Upload it and ask questions in plain English. Want to update a column based on a complex rule? Just say so.

The key differentiator from tools like ChatGPT’s code interpreter is that BitBoard preserves direct cell-level control. You are not just getting an output. You are getting a live spreadsheet that you can inspect, modify, and extend. The AI builds the structure, but you own it. This is an important distinction because finance and operations people need to audit the work. They need to see the formulas, check the logic, and make manual adjustments. A black-box output is not useful in a world where the CFO needs to explain the assumptions to the board.

The company claims it handles complex modeling and analysis ten times faster than traditional spreadsheets. That is a bold claim, but it is directionally right if you compare “describe what you want in a sentence” to “spend an hour building nested IF statements and INDEX MATCH formulas.”

Two-person team. San Francisco-based. The Palantir and BlackRock backgrounds are relevant here. Both companies live and die by complex data analysis, and the founders have seen firsthand how much time smart people waste fighting with spreadsheets instead of thinking about the data.

The Verdict

I think BitBoard is chasing the right insight: the spreadsheet interface is overdue for reinvention, and conversation is probably the right paradigm shift. The people who would benefit most from powerful spreadsheet capabilities are often the ones who do not have the technical skills to use them. A chat-based interface collapses that gap.

The cell-level control is the detail that matters most. Every other AI data tool produces outputs. BitBoard produces editable spreadsheets. That is the difference between a calculator and a tool. Finance teams, operations managers, and analysts need to trust, verify, and modify their models. If the AI just hands you a number, you cannot use it for anything serious.

The competitive landscape is crowded though. Rows, Equals, and Airtable all have funding and traction. Excel and Google Sheets are building their own AI features. Copilot in Excel already writes formulas and generates insights. BitBoard needs to be meaningfully better than “Excel plus Copilot,” which is a high bar because Copilot is free for enterprise Office 365 users.

The positioning matters here. If BitBoard targets the same use cases as Excel, they lose. Excel has two billion users and infinite distribution. But if BitBoard targets the use cases that are too complex for Excel but not complex enough for Palantir, there is a real market. Think mid-market finance teams building models that are currently held together by duct tape and prayers. Think operations teams that need to analyze data from five different sources but do not have a data engineering team.

Thirty days, I want to see the user retention curve. Spreadsheet tools have notoriously high trial-to-churn rates because people try them for one task and go back to Excel. Sixty days, whether teams adopt it or just individuals. B2B spreadsheet tools need team adoption to generate meaningful revenue. Ninety days, the integration story. Can BitBoard pull data from the tools that feed spreadsheets today, like Salesforce, Stripe, QuickBooks, and HubSpot? The standalone spreadsheet is a demo. The connected spreadsheet is a product. The founders have the right pedigree and the right instinct. Now they need to find the wedge that makes people switch.