Obsidian: The Anti-VC, Anti-Lock-In Knowledge System Built by Builders

To define the subject clearly: the Obsidian discussed here is the note-taking / knowledge management app built on local Markdown files, emphasizing bidirectional links, plugins, and data ownership—not the game studio. The public About page identifies the two core founders as Erica Xu and Shida Li; as of 2026, Shida is CTO, Erica is COO, and the CEO is Steph Ango, who joined later. Legally, the operating entity is not named Obsidian but Dynalist Inc., which is explicitly stated in the privacy policy, terms of service, and license overview.

The part many people want to know—but where public information is thinnest—is family background. Public materials are limited on both founders’ birth years, parents, household class, early family resources, and siblings. Only a few fragments are confirmable: a 2015 Quizlet intern introduction described Erica as “originally from Nanjing, China,” while Shida’s publicly accessible professional traces confirm that he developed his career in Canada, but his birthplace, parents, and early household environment cannot be confirmed from open sources. In other words, both founders have been notably restrained in disclosing private life details.

What is much easier to see than their “family background” is their “way of building.” Erica’s personal site states that she is passionate about education, cooking, design, programming, and cats; she openly criticizes the current education system as outdated and has long pursued projects framed around first-principles thinking. This matters because it helps explain why Obsidian was never designed as a workflow that dictates how users must work, but as a toolset that lets users construct their own system. Public information on Shida is sparser, but his published comments on community-building show that he brought prior gaming-community moderation experience, which directly shaped Obsidian’s choice to grow through community and product depth rather than through sales-led expansion.

On education, the clearest through-line is that both founders met and collaborated deeply at the University of Waterloo. Erica’s public profile explicitly shows a 2011–2016 Computer Science bachelor’s degree there. For Shida, University of Waterloo news confirms that he was a second-year Software Engineering student in 2013, and public LinkedIn snippets also point to a Software Engineering background. Erica also completed at least one clearly visible internship while in school: in 2015 she was a Software Engineer Intern at Quizlet. Whether and when Shida completed his degree, and their full co-op histories, remain public-materials limited.

Their collaboration clearly predates Obsidian by years. Erica’s personal website lists multiple joint projects with Shida, including MyFood, UW Course Indexer, and Marmo UI, showing that they had already been repeatedly building practical tools together before Obsidian. In parallel, University of Waterloo also recorded Shida’s 2013 win at a Nokia imaging hackathon with a “Smart Resizing” image app. Put together, these pieces suggest a very consistent pattern: they did not start with a company and then go looking for a product; they first formed a long-term habit of making tools together, and only later turned the strongest of those tools into sustainable businesses.

If “first representative work experience” means the earliest public point at which they clearly appear as professional builders, Erica’s record is easier to trace. She had a Quizlet internship, and the outdated résumé section of her personal site preserves a very strong testimonial from a MappedIn manager, showing substantial execution ability in product and interface work even as a student. Shida’s conventional employment history is much less public, but his visible path is more that of a technical builder: hackathon work, GitHub repositories from the mid-2010s, and later responsibility for infrastructure and performance in Dynalist and Obsidian. This suggests that their division of labor formed early: Erica leaning more toward product definition, design, operations, and outward articulation; Shida leaning more toward implementation, performance, infrastructure, and overall technical direction.

The first major bridge into their core field was Dynalist. Erica’s site states that Dynalist began in March 2015 and aimed to become the best cross-platform outliner. A University of Waterloo capstone page then shows that a “Team Dynalist” project had already made over US$13,000 in the preceding few months by January 2017, was profitable, and growing fast. A key nuance matters here: the school page presents Dynalist as a five-person capstone team, while the long-term public founder attribution centers mainly on Erica and Shida. The safest synthesis is that Dynalist appeared in an academic setting as a multi-person capstone team, but the ongoing public entrepreneurial attribution later settled primarily on Erica Xu and Shida Li; the later commercial roles of the other listed teammates cannot be confirmed from public materials.

Obsidian was not born out of formal market research. It came from a classic founder pattern: long-term dissatisfaction with existing tools, followed by finally deciding to build the desired product themselves. Erica said in interview that she had carried the basic idea for about two years before they started working on it, and that it was during the 2020 quarantine period that they finally executed. She had tried building personal wiki systems before and remained dissatisfied with existing note apps because there was always some crucial detail missing or implemented in the wrong way. Founders of this kind often build unusually sticky products because they are solving a persistent, embodied problem of their own rather than abstracting for an “average user.” That is why Obsidian feels less like a general-purpose collaboration layer and more like a personal knowledge environment.

Two early product decisions were especially decisive. First, the plugin architecture was not retrofitted later; it was designed in from day one, and even many native features were implemented as plugins. Second, product testing and growth were not built around closed PM loops but around Discord private beta and, later, a forum, allowing product development, community governance, and plugin ecology to grow together. Shida publicly explained that the plugin marketplace had to live inside the product, because they did not want discovery, installation, trust, and safety to be left to a fragmented external ecosystem. That decision later became one of Obsidian’s strongest sources of network effects.

In terms of brands, assets, organizations, and platforms, Obsidian is not the kind of company that accumulates many acquired assets under a large parent structure. Instead, it concentrates on a small number of tightly integrated assets. These include the desktop and mobile apps, the local vault file model, official plugins and developer documentation, the community plugin/theme marketplace, Obsidian Sync, Obsidian Publish, Catalyst, optional Commercial licenses, the Enterprise page, the forum and Discord community, and later the Web Clipper. Two especially important “influence assets” stand out: the public manifesto/help system, and open-format/specification work such as JSON Canvas. The latter may not monetize directly at once, but it greatly strengthens the brand promise of openness, durability, and non-lock-in.

The revenue model is not complicated, but it is highly disciplined. The core app is free, requires no mandatory sign-up, includes no ads, and the company says it collects no app telemetry; paid revenue comes mainly from official add-on services and support-style licenses. The official pricing pages show that Sync has Standard and Plus tiers starting at US$4/month annually, Publish starts at US$8/month annually, Catalyst is a one-time support license starting at US$25, and the Commercial license is now optional rather than mandatory. Structurally, this is “free core tool + paid convenience layer + support layer,” not “free until lock-in, then monetize switching costs.” The company is selling a smoother official service layer, not compulsory control over user data.

That model did not begin with Obsidian. The Dynalist era already shows a freemium / Pro pattern: the public page visibly distinguishes free use from Dynalist Pro, and premium features include uploads, unlimited bookmarks, mind map view, recurring dates, and more. In other words, the founders had already tested a “free base + paid advanced features” model before Obsidian. The 2017 university record of early Dynalist revenue shows that they learned quite early how to make a niche productivity tool into a real, survivable business. Obsidian then evolved the model further: away from simply charging to unlock more features, and toward charging for better official services and for long-term independence itself. That is the key shift in the business model’s evolution.

On capital and partnerships, the clearest conclusion is negative: the company repeatedly states that it has no investors and is 100% user-supported. The official manifesto, Steph Ango’s 2023 CEO-joining post, and later pricing and licensing changes all reinforce the same message. That means Obsidian does not sit inside the usual “fund–board–growth target–exit” startup structure. Its real long-term resource network is closer to “user payments + community contribution + small team iteration.” The key durable nodes in that network are threefold: the founders’ own long-running partnership, the moderators/plugin authors/theme makers/users who became part of the company’s operating strength, and later additions like Steph Ango, who first entered as a power user and contributor before becoming management.

At least six turning points matter most. First, launching Dynalist in 2015 validated that the pair could collaborate to build a niche productivity product with real revenue. Second, 2020 quarantine turned Obsidian from idea into execution. Third, the day-one plugin architecture ensured that Obsidian would become an ecosystem substrate instead of a closed note app. Fourth, the 2022–2023 transition around version 1.0 and the addition of Steph Ango as CEO created a new organizational balance, allowing the founders to spend more time on what they each did best. Fifth, 2024’s JSON Canvas work and repeated third-party security audits turned “open” and “secure” from slogans into more institutionalized assets. Sixth, the 2025 decision to make work use free and commercial licenses optional pushed the company even further toward “support by alignment” rather than “control by licensing.”

Obsidian’s greatest achievement is not merely that “another note app became successful.” Its deeper achievement is that it translated a once relatively niche, geek-heavy knowledge-management worldview into a mass-accessible product form. By 2023, media coverage relayed the company’s own estimate of about one million users, more than 110,000 Discord members, and more than 94,000 Reddit members; official pages consistently described the ecosystem as having thousands of plugins and themes; and the 2023 Gems of the Year post alone recorded 287 nominated community projects. By 2025, the company said that more than 10,000 organizations were using Obsidian for work. For such a small team to push local files, linking, plugins, and personal knowledge bases to that level is not just product success; it is narrative success.

Why does the outside world remember it? There are four layers. First, it openly aligns with “your files are yours,” “open formats first,” and “no lock-in,” which makes it feel strongly countercultural in a cloud-first era and helps build loyalty. Second, it reframed note-taking as note-making: not just recording information, but structuring personal knowledge. Third, it did not treat community as outsourced support; it treated community as product capability—plugins, themes, tutorials, competitions, moderators, forums, and even hiring from the user base all became one coherent system. Fourth, it proved that a very small team could build and maintain a sophisticated cross-platform app for years without venture capital and without advertising. What people remember is not one feature, but a whole method of organizing product, company, and ideology.

The criticisms are real, but they are mostly ideological and product-direction criticisms rather than major public scandal. Five themes recur. First, the closed-source debate: since 2020, users have repeatedly pushed for open sourcing, and the founders publicly argued that small-team open-source stewardship would be too expensive, that open source does not automatically make software safer, and that they preferred privacy statements, optional zero-network use, and third-party security audits. Second, plugin security: the company itself warns users that community plugins require trust judgments, and Restricted Mode is on by default, which means ecosystem growth necessarily widens supply-chain risk. Third, the learning curve: Markdown, configuration depth, and plugin dependence remain barriers for new users. Fourth, pricing debates: especially around Sync / Publish and older commercial license logic. Fifth, culture debates: one plugin developer described parts of the community as “Markdown maximalism,” meaning some long-time users may over-prioritize text purity and structure in ways that make the product feel intimidating or mildly gatekept to newcomers. Publicly visible controversy is concentrated in these areas.

Obsidian’s current real-world position can be summarized this way: it has moved from being a pandemic-era hacker’s tool to being a small, hard, high-conviction software company with a strong community and a strong data-sovereignty narrative. In 2025 the company said it was being used in more than 10,000 organizations; the official Enterprise page lists support and usage from large institutions including Amazon and Google; Steph said in a 2025 interview that the company then had only seven full-time employees; and the 2026 About page still shows a very small human team. That means Obsidian is not in the position of “the biggest platform company.” It is in the position of “one of the strongest independent paradigm examples in productivity software.” The parts that still cannot be seen clearly in one pass are the founders’ birth years and family details, Dynalist Inc.’s equity structure, real revenue and profit, valuation, Shida’s full education / early-career record, and the exact downstream role of other members listed on the 2017 Dynalist capstone page. On those points, public materials remain limited / sources differ / confirmation is not currently possible.