When Microsoft disclosed in late January that only 15 million of its 450 million-plus commercial Microsoft 365 users were actually paying for Copilot, the implications rippled beyond a disappointing 3 percent adoption rate. It suggested something more fundamental: enterprises want AI-powered productivity tools, sure, but they're not finding what they need in Microsoft's flagship offering.
Enter o11—a four-person startup that emerged from Y Combinator's Winter 2026 batch with a pitch so direct it borders on audacious. "Microsoft Copilot, But It Actually Works," reads the tagline. It's the kind of positioning that makes venture capitalists wince and competitors take notice in equal measure.
Co-founders Aryah Oztanir and Ajay Misra aren't betting against workplace AI. They're betting against the sidebar. And in the process, they're testing a provocative thesis: that Microsoft's tepid Copilot numbers don't reflect skepticism about AI assistants, but frustration with how those assistants actually show up in daily workflows.
Working Where Work Actually Happens
What makes o11 different—at least according to the founding team—is that it doesn't live in a chat window. The product operates as a suite of AI-native add-ins running directly inside PowerPoint, Excel, and Word. Users generate slides, build financial models, draft long-form documents without ever leaving the canvas where work gets done.
The company's Y Combinator profile calls it "the first full-featured AI-native execution layer within the entire Microsoft 365 suite." Whether that's technically accurate depends on how narrowly you define "full-featured," but the intent is clear enough.
In PowerPoint, the add-in can spin up brand-compliant decks in seconds. In Excel, it automates model construction and includes what o11 calls "Audit Mode"—a feature designed to catch linking errors, semantic inconsistencies, logical gaps. The Word add-in handles longer documents with tone control and citations that trace back to specific source paragraphs.
Then there's the charting piece. In mid-February, the team shipped natural-language chart generation for the complex enterprise visuals that typically require specialized tools: Gantt charts, waterfall diagrams, Mekko charts, 3D visualizations. "McKinsey-level, editable charts in PowerPoint," Oztanir wrote on LinkedIn. It's a not-so-subtle jab at incumbents like think-cell, which has dominated consulting workflows for the better part of two decades.
The Context Play
The other core technical claim is what o11 calls "infinite context"—no limits, the company says, on how many files or documents it can analyze simultaneously. Every output comes with full citations linking back to source paragraphs.
For investment bankers and capital markets analysts—the startup's early target audience—that's a big deal. The enterprise landing page highlights integrations with virtual data rooms (Datasite, Intralinks) and financial data platforms (PitchBook, S&P Capital IQ, FactSet). The pitch is straightforward: pull from dozens or hundreds of documents at once, generate models or decks on the fly, and trace every assumption back to its origin.
Whether the system truly handles "infinite" context at scale, or how it stacks up against competitors in real-world stress tests, remains to be independently verified. As a positioning move against tools with visible file caps, though, it works.
Four Tiers, One Strategy
Pricing runs from free to enterprise custom, with two paid tiers in between: Pro at $20 monthly, MAX at $200. The ladder suggests a dual strategy—prosumer on-ramp meets high-value enterprise push.
For larger organizations, o11 promises single sign-on through Okta or Azure AD, SOC 2 compliance (currently in process), and deployment flexibility across SaaS, private cloud, or on-premises environments. Distribution happens through standard Office add-in installation, though no public AppSource listing appeared in searches for this piece. That could signal enterprise-only deployment for now, or side-loading while the team builds toward broader availability.
The go-to-market has been scrappy and social. Oztanir and Misra ran a LinkedIn campaign in mid-February offering "300 FREE credits" tied to post engagement. Their Product Hunt launch on January 14 drew generally favorable comments, particularly around Excel and PowerPoint utility. The company claims users "around the world" but hasn't publicly named customers. A logo strip on the website—Yale, Mayo Clinic, Bank of America, Wells Fargo—appears to reflect founder pedigree rather than client endorsements, though that distinction isn't spelled out.
A Crowded Field
The competitive landscape is dense. Think-cell remains the enterprise standard for consulting-grade charts. UpSlide offers AI-powered consistency checks, Excel-to-PowerPoint linking, brand management. Empower Charts and Formato Charts serve similar niches. All come with entrenched user bases and years of iteration.
Microsoft isn't standing still either. Copilot continues to evolve, and the Office Add-ins platform is moving toward unified manifests and what Microsoft calls "agents with add-in actions"—basically, tighter cross-app integrations. That could help a tool like o11, which spans multiple applications. It also means Microsoft could close its own gaps, or that larger players could match o11's feature set over time.
In a company blog post, o11 contrasts itself with search-oriented platforms like Hebbia and Rogo, calling itself "the only platform that runs natively across Excel, Word, and PowerPoint seamlessly." Like most startup claims, independent validation is pending.
The Underlying Wager
At its core, this is a bet that the Microsoft 365 installed base—450 million commercial seats and counting—represents a vastly underserved market for AI tools that genuinely integrate with knowledge work. If Copilot adoption is stalling at 3 percent, the reasoning goes, there's space for alternatives that nail specific pain points: complex charting, multi-file reasoning, in-canvas editing.
Y Combinator partner Jared Friedman endorsed the "Copilot, but it actually works" framing in a LinkedIn post recounting the team's interview. The startup raised $500,000 from YC and is now live. Four people, a confrontational tagline, a product aimed directly at professionals who spend their days inside Office apps.
Whether o11 can convert that opportunity into sustainable revenue will hinge on execution, distribution muscle, and whether the product delivers on its technical promises when usage scales. For now, it's another data point in a broader story: the race to build genuinely useful workplace AI is accelerating, and startups are perfectly willing to challenge incumbents when the door gets left open.
Even just a crack.
