Altrina Wants to Record Your Workflow Once and Never Make You Do It Again

The Macro: Everyone Has SOPs, Nobody Wants to Follow Them

Standard operating procedures exist in every organization above about five people. They live in Google Docs nobody reads, in Notion pages that are always six months out of date, in the heads of the one person who knows how things actually work. When that person goes on vacation, everything slows down. When they leave, things break.

The automation market has been trying to solve this for years. Zapier connects apps. UiPath builds bots. Make (formerly Integromat) lets you draw flowcharts. They all work, to varying degrees, but they all require you to translate your actual workflow into their abstraction layer. You have to think about triggers and actions and conditional logic. For simple stuff, that is fine. For the messy, multi-step processes that actually eat hours out of someone’s day, it is a project unto itself.

The RPA (robotic process automation) market hit roughly $2.9 billion in 2024 and keeps growing, which tells you two things: businesses desperately want to automate repetitive work, and the existing tools have not fully solved the problem. If they had, the market would be consolidating, not expanding.

What is shifting now is the interface. The old model was “describe your workflow in our language.” The new model, which several companies are chasing, is “show us your workflow and we will figure it out.” That is a meaningful difference because it moves the skill requirement from the tool to the task. You do not need to know how automation works. You just need to know how to do your job.

The Micro: Show, Don’t Code

Altrina’s core product, Altrina Automator, works by watching you do something in your browser and then replicating it. Record yourself logging into a portal, downloading a report, reformatting it, and emailing it to three people. Altrina captures that sequence and turns it into a repeatable automation that you can edit using plain English. No code, no flowcharts, no “if this then that” logic trees.

Mo Nasir and Harvey Hu founded the company out of Menlo Park. Mo previously worked on self-driving cars at NVIDIA, and Harvey was a machine learning engineer at Google and TikTok. That background matters here because browser automation at this level is essentially a perception and planning problem. The system has to understand what it is looking at on screen, decide what to click, handle variations in page layout, and recover when something unexpected happens. Those are the same categories of problems you solve in autonomous driving, just at lower stakes.

The company came through YC’s Winter 2025 batch and has positioned itself across a broad set of industries: local governments, law offices, insurance brokers, consulting firms, hedge funds. That breadth is ambitious for an early-stage company, but it makes sense if the core technology genuinely generalizes. A browser automation that works is a browser automation that works, regardless of whether you are filing a compliance form or reconciling invoices.

The numbers they share publicly are solid. Over 89,000 processes completed, and clients reporting 30 to 100 percent workflow automation with 20 to 90 hours saved weekly. They have also picked up SOC 2 Type I and II, GDPR, HIPAA, and ISO 27001 certifications, which is an unusually heavy compliance portfolio for a company this young. That signals they are serious about enterprise sales, where security reviews kill deals faster than bad demos.

The competitive landscape is crowded. Zapier and Make own the “connect two apps” tier. UiPath and Automation Anywhere dominate traditional RPA. Newer entrants like Bardeen and Browse AI are also chasing the browser automation angle. What Altrina is betting on is that the “record and replay” paradigm, combined with natural language editing, is a better on-ramp than any of those alternatives offer. I think that bet has legs, especially for the mid-market companies that are too small for UiPath but too complex for Zapier.

They offer a free tier with $10 in credits to start, which is the right move for a product that needs you to experience it to understand the value.

The Verdict

I like the approach. The “show me once” model for workflow automation is more intuitive than anything the established players offer, and the founding team has the technical depth to make the underlying computer vision and planning systems work reliably. The compliance certifications suggest they are building for real enterprise buyers, not just demo-friendly prototypes.

The risk is the same one every horizontal platform faces: trying to serve law offices and hedge funds and local governments simultaneously is a lot of context to carry. The best automation tools tend to get really good at one domain before expanding. If Altrina can nail reliability, which their WebVoyager and REAL benchmark scores suggest they are working on seriously, the breadth could be a strength. If the automations break in edge cases that vary by industry, the breadth becomes a liability.

In 30 days, I would want to see how many recorded automations actually run successfully on the tenth attempt without human intervention. In 60 days, I would want to know whether customers are building new automations themselves or relying on Altrina’s team. In 90 days, the question is whether any single vertical is pulling away as the dominant use case. The product is real and the team is strong. The execution question is whether “works everywhere” can actually mean “works everywhere.”