Shannon Wu Wants the Computer to Let You Get On With Your Life

Shannon Wu would like you to stop staring at your computer quite so much. This is an unusual sales pitch from someone building an AI company. Technology tends to arrive with an appetite for attention and a confident explanation of why it deserves another helping. Wu’s proposition begins with the objects already around us. You should be able to speak to them, get something done, and carry on with your day.

She is the founder and CEO of OpenHome, which gives developers software and hardware for building conversational devices. The speaker is a starting point. The larger ambition includes toys, appliances, and other objects that can respond to ordinary speech. A room, in this account of computing, has possibilities beyond being somewhere to charge your phone.

Wu arrived here through growth marketing, digital identity, and decentralized finance. Those subjects do not immediately suggest microphones. They do involve a recurring practical problem: a technology can be available long before people know what to do with it. Her career has repeatedly occupied that awkward interval between a technical idea and its place in someone’s life.

The Stanford friends who kept building

In 2012, while studying at Stanford, Wu learned about blockchain. She later recalled becoming absorbed in it, along with a close circle of friends. Years afterward, that circle would work together on Bloom. The connection matters because it gives a human scale to a career usually described through company names. Before the organizations, there were people sharing an interest.

In a November 2017 introduction to Bloom, Wu named fellow Stanford connections including John, Jesse, Alain, Devon, and Daniel. Jesse Leimgruber would later become OpenHome’s cofounder and CTO. Their names therefore recur across two quite different projects. A technology cycle may have a short memory; people who have already built together can have a longer one.

After graduating, Wu founded Mr. Progress, a growth marketing consultancy working with emerging technologies. Her work included bringing products in automation, AI, data, and blockchain to market. She also became involved with MakerDAO. Growth work requires a particular kind of translation: explaining enough of the machinery for someone else to see a reason to use it.

Before founding Mr. Progress, Wu worked on the investment team at FOUNDER.org, an early-stage fund. There she supported more than 40 founders through a program that combined investment, executive mentors, industry partners, and a curriculum. It was a view of company building from the supporting side of the table. A founder needs resources, but also introductions, feedback, and a way to put the next decision in order. That earlier work belongs beside the consultancy in her story: both involved helping other people turn a promising start into a functioning business.

At Bloom, she worked on the founding team and led go-to-market strategy. Its subject was digital identity: how people could prove something about themselves while retaining control over the information they shared. That is less cinematic than a talking device. It is also intimate. Identity systems sit between an individual and the institutions asking to be trusted with their details.

“For me, the most important mission is for people to take back control of their data.”

Shannon Wu, 2019

Explaining the machinery

Wu’s work extended beyond a launch. Bloom’s author archive lists 120 posts under her name, covering product integrations, digital signatures, data breaches, and economic research. It is a record of the explaining that follows the announcement. Once a technology exists, someone still has to describe its uses, answer questions, and give the next person a place to begin.

The subjects in that archive are concrete. A digital signature needs to be checked. An employee wants an expense reimbursed. A person wants to know whether their information has been exposed. There is a useful discipline in those examples. An ambitious technical architecture eventually has to meet a task small enough to explain in a sentence.

She also cofounded Identity Review, a digital identity consortium. Her work at Bloom included growth, government relations, and financial institutions. It is a reminder that the audience for a technology includes organizations with procedures, obligations, and questions that cannot be settled by an enthusiastic demonstration.

Teaching appears in her career, too. She has guest lectured on marketing and search optimization for Stanford’s E145 technology entrepreneurship class. Former collaborators have described her support for students. The visible work is practical: helping people understand a subject well enough to act on it. A good explanation earns its keep when the listener can do something afterward.

A company with room for other people’s ideas

OpenHome puts Wu’s work into physical objects. Its platform includes an operating system, voice software, development hardware, and tools for building AI agents. The company presents local operation and privacy as part of that approach. Those are product commitments with practical consequences: where processing happens, what a developer can change, and how much a device depends on a remote service.

For Wu, the developer community is central. She has connected her experience around MakerDAO with her approach to OpenHome, describing outside builders as a source of applications she would not have imagined herself. There is a modest admission inside that strategy. The people running a platform do not possess a complete inventory of what it might be good for.

In June 2024, she authored the announcement of OpenHome’s developer grant program. It offered selected projects up to $50,000, with speakers, toys, and appliances among the proposed applications. Selection involved more than describing a promising idea. The announcement emphasized earlier projects, independent execution, community participation, and evidence that a builder could deliver.

The milestones made that expectation tangible: share a repository, produce a working demonstration with open code, and integrate the project through review and testing. The invitation came with homework. That is a reasonable bargain for a platform whose usefulness depends on what other people contribute. A development kit becomes interesting when someone makes it do something particular.

Two sisters, two companies

There is another founder in Wu’s family. Her sister Amy founded Manifest, and in January 2025 the two joined Tom Wallace and Saxon Baum for the Skin in the Game podcast. They discussed their separate entrepreneurial paths, their connection to Stanford, and support from Florida Funders. The family relationship is a small, specific detail in a professional life otherwise crowded with platforms.

The appearance put two company builders in the same conversation without collapsing their work into one venture. Both discussed what small teams could accomplish with AI. Their account also emphasized the value of investors who offered personal attention. For a founder, there is a difference between having access to capital and having someone willing to engage with the particulars of the company.

The sibling connection adds a different kind of continuity to Wu’s story. Her early account of Bloom begins with friends; this conversation includes a sister building her own business. Her career contains plenty of technical nouns. It also contains people returning to one another to compare experiences, offer support, and work through what comes next.

The host learns to listen

By October 2025, Wu had a conversation series of her own. OpenHomies began with Ahmed Bouzid, the CEO of Witlingo and a former product leader for Amazon Echo Smart Home. Inviting someone with experience inside an earlier generation of voice products gives the discussion a useful perspective. A new interface still has to live with the habits of the people using it.

The episode covered the distinction between issuing a command and holding a conversation. It also addressed echo cancellation and sensitivity to pauses. These are wonderfully unglamorous details. A pause may mean that someone has finished speaking, or merely that they are looking for a word. Getting that wrong can turn a supposedly helpful object into an impatient dinner guest.

For Wu, hosting makes another part of the work visible: asking experienced practitioners about the decisions underneath a product. The discussion included what earlier voice platforms learned about their users and the design principles behind a more comfortable interaction. Conversation is easy to advertise. Making room for the other person to finish a sentence is a more exacting standard.

Why the next conversation might be in Japanese

Japan gives OpenHome’s ambitions a particular setting. Wu has pointed to the country’s consumer hardware history as a reason to work with its builders. In July 2026, OpenHome announced a selective developer program with ElevenLabs. Accepted participants could receive voice credits, with documentation and community support in Japanese and English. The company said it had been shipping development kits there since spring.

The applications it described included a companion that reads and responds to YouTube Live chat and an ambient task manager. These are individual jobs for a voice interface, each with a recognizable reason to exist. The program also acknowledged how demanding Japanese speech can be, with pitch accent, politeness, and intonation affecting the experience. Fluent conversation involves much more than pronouncing the words.

Wu’s stated ambition reaches people of different ages and languages, with less time tethered to a screen. That remains an ambition, and the work has to survive ordinary use: an interruption, an unexpected request, a device someone wants to customize. Her career has supplied experience in explaining technology and inviting others to build with it. OpenHome now places that work within speaking distance.

There is a pleasingly everyday test waiting at the end of all this. Someone asks a device for help. It understands, responds, and lets the person get back to whatever they were doing. For a founder building talking computers, a successful moment might be the one when the computer has the courtesy to stop talking.