We're redefining "zero-trust" for the agentic era.
Joined February 2018
- Tweets132
- Following90
- Followers351
- Likes170
We cut Multifactor down to one thing: sharing an online account by sending a link.
Pick what each link can do, watch every action live, revoke in one click. Your password never leaves.
Built in our second Lock-In Week. Try it (Hacker News for now): lite.multifactor.com
multifactor.com/blog/introdu…
Cloudflare Workers are fast and scale on their own, but nothing lets a third party check what code a Worker is running.
Introducing Notary. Notary pairs a Worker with two TEEs so anyone can cryptographically verify the code that is live on a domain.
multifactor.com/blog/introdu…
An AI agent just retrieving an Amazon invoice has access to an account that can do much more.
We explored how to map those actions, covering surface discovery, request classification, and parallel browser workers, to keep agents safely in their lane.
multifactor.com/blog/mapping…
Do you share any online accounts with other people or AI agents? Partner, teammate, contractor, an agent that logs in for you — personal or work.
How do you do it today, and what do you wish was better?
Collecting stories. Replies/DMs welcome.
33%I share accts w/ people
0%I share accts w/ agents
67%Both
0%Neither
3 votes • Final resultsYou gave your agent access to your X account and asked it to summarize your DMs.
How do you know it read all of them? How do you know it didn't also reply to posts, or make its own?
Password managers handle authentication, not authorization. OAuth and MCP need the website to play along, and your regional bank isn't shipping an MCP server.
We built Checkpoint: scoped access to any account, enforced at the network layer inside a TEE, delivered as a hyperlink anything can open.
multifactor.com/blog/checkpo…
We cleared four days, cancelled every meeting, gave five engineers unlimited agent credits, and made everyone demo live on Thursday.
multifactor.com/blog/lock-in…
Multifactor (YC F25) retweeted
Doing a survey for something we're working on at @MultifactorCOM:
-What online accounts are you most often sharing with friends/family/significant others?
-Are there any online accounts you've wanted to share with them but haven't felt comfortable doing it?
Reply or DM me!
We hired Multifactor's founding engineering team in 6 weeks — and we didn't start with a job post. We started with a rubric.
Six pillars, no take-home coding tests, live paired sessions where AI tools were encouraged (we tested judgment, not whether you could prompt an LLM), and finalists flown to SF for a day of whiteboard design + culture fit.
The story: multifactor.com/blog/how-we-…
The line between design and engineering was never about skill.
It was about who could act on the design directly.
AI tooling moved that line. New post on becoming a design engineer at @Multifactor: multifactor.com/blog/becomin…
For an AI agent to do real work (pay an invoice, pull a statement, check an order), it has to get through a login page first. Login pages are built for humans: multi-step forms, one-time codes, passkeys, captchas, layouts that change overnight.
Neither naive approach held up for us. Scripts break on every redesign. And letting an AI improvise with real credentials on a live account is not something we were willing to ship.
At Multifactor, we split the problem:
- An AI recon agent explores each website ahead of time, with no customer credentials involved, and produces a deterministic, auditable login flow.
- Every flow is tested in our lab with our own agent credentials, then manually reviewed and released individually. When a site changes overnight, the flow self-heals in the lab and re-enters that same review gate.
- At login, Checkpoint replays that flow inside a trusted execution environment. No AI in the loop when your credentials are in play.
- Success is verified with evidence: signals the login itself caused, instead of waiting and hoping.
Full write-up by our founding engineer Pierre Ortega: multifactor.com/blog/teachin…
Together: no point where a usable secret sits anywhere a thief or rogue agent can reach it. That's Checkpoint — account access that's scoped, revocable, and auditable.
Dave Blumenfeld tells the full story:
multifactor.com/blog/keypo-m…
Two teams solved opposite halves of this problem:
- Keypo -> let an agent USE a credential without ever seeing it (secure enclaves)
- Multifactor -> protect that credential at rest with no single master-password point of failure (MFKDF)
An AI agent is only as useful as the access you give it. But the more access you grant, the more damage it can do.
The credential that checks your balance is the same one that drains your account. So how do you share an account with an agent safely?
Multifactor (YC F25) retweeted
Making a strategic acquisition as a seed-stage company was not a decision I took lightly, but so far all I'm wondering is why we didn't do this sooner.
We first met the Keypo folks about 6 months ago during the YC batch, and were blown away by not only their alignment on highly complementary problems in the agent security space, but more importantly the strong common principles and values that underpin both of our companies.
Behind the scenes, we've been working with Dave full-time for about a month while the acquisition was being finalized, and immediately noticed a boost in energy and productivity across our entire team.
Dave's background and experience in security, authentication, AI agents, and trusted hardware is unparalleled. He's been an absolute pleasure to work with so far, and I'm so excited to share updates about what we're building together in the coming weeks and months.
Learn more about Dave and our acquisition of Keypo: multifactor.com/blog/keypo-a…
6 months ago, we met a team that was remarkably well aligned with our agentic security vision.
Today, we're joining forces.
We've officially acquired Keypo, Inc., and its founder Dave Blumenfeld is joining us as Founding Engineer, bringing his work on secure enclaves directly into our platform.
Multifactor (YC F25) retweeted
We share our offices with @MultifactorCOM, one of the best AI agent security companies out of YC.
Our claude has context that we do this. So, when asked, Opus 4.6 created this video for Multifactor.
Passwords are so last epoch with Multifactor.
Used this prompt with some modifications to create a YTP video for my startup.
This is the first piece of pure AI slop that is good. And it's because it had all my context on my company.
@KoyalAI is an agentic AI filmmaking platform that converts scripts into films by intelligently routing across 40+ AI video models to get the best output for each shot or style.
Please don't say AI doesn't understand creativity now. Yes, we should be afraid but dw, we're making sure it helps humans.
The best use of AI is with human creativity. But this is very close lol.
Multifactor is hiring an Engineering Lead!
Major companies like Cloudflare, Browserbase, Vercel, and Firecrawl have called agent auth one of their biggest unsolved problems.
Multifactor has a validated and differentiated solution, and we need your help to ship it faster.
Apply now: ycombinator.com/companies/mu…
Multifactor (YC F25) retweeted
Congratulations to Multifactor on the $15m seed round straight out of YC's fall batch.
This was one of our earliest and fastest investments in F25 at @pioneer_fund. Outlier founders.
They're making it possible for AI agents to operate accounts securely.
Multifactor (YC F25) retweeted
We're proud to be leading $15M seed round into @MultifactorCOM, to make online accounts fundamentally safe for the age of AI agents. Securely connecting our accounts to AI tools, while keeping users in control, is the need of the hour.
As AI agents become part of how we collaborate, the need for mathematically provable identity foundations has never been clearer. Multifactor is building exactly that - a secure way for humans and AI to share access without exposing credentials.
Vivek and @Autoparallel - Thanks for picking @NexusVP as your partner! Excited to be on this journey with you, along with an incredible group of supporters and advisors, including @Padmasree, @gokulr, @collinmathilde, @auren, @ycombinator, @dessaigne, @Liquid2V, and more.
As Vivek mentions in his post: Identity should be our strongest defense, not our biggest vulnerability. Onward!
See @eringriffith's feature of Multifactor in today's @nytimes article: nytimes.com/2025/12/04/techn…
#AI #Identity #Security
