The visa problem that made Alma change its business

Aizada Marat knew her way around law firms. She had trained at Harvard Law and worked at Cooley. Yet when she needed her own immigration status sorted, legal advice failed her badly enough to interrupt her ability to work. Expertise, it turns out, is a poor substitute for a well-run process when the process is happening to you.

Marat later met Assel Tuleubayeva and Shuo Chen, and the three built Alma. At first, the plan sounded like the usual legal technology pitch: give immigration lawyers software to make their work faster. The founders raised an initial $500,000 to develop a minimum viable product. Early conversations with lawyers confirmed that repetitive work was a problem. But as the team built and talked to customers, a less convenient truth emerged. The client was buying an outcome and a relationship, not a faster form.

The first product did not reach the real customer

The original idea was sensible. Immigration petitions involve recurring questionnaires, form fields, document requests, deadlines, and evidence files. Software can save attorneys from copying the same detail into the same box yet again. Marat had seen that productivity gap in conversations with law firms. But a more efficient lawyer’s desk would not, by itself, tell an applicant whether an O-1 or EB-1A made sense, what evidence was missing, or why a case had gone quiet.

Marat’s own account of the pivot is unusually plain: “Immigration wasn’t just a legal workflow problem; it was an end-to-end services problem.” That sentence changed the company’s customer. Alma moved from building a tool for law firms to offering immigration support to individuals and businesses. Chen, an engineer who had worked on machine learning at Uber, joined as co-founder and CTO. The founders say they have held about fifteen visas among them. It is hard to regard a deadline as an abstract workflow when you have lived under one.

“Immigration wasn’t just a legal workflow problem; it was an end-to-end services problem.”Aizada Marat, co-founder and CEO

There is an important distinction behind the brand. Almaca, Inc., doing business as Alma, operates the platform. Legal services come from Alma Legal Services, P.C., an independent law firm, under a separate attorney-client agreement. Alma’s intake can use AI assistance, but its materials say an attorney reviews the output and makes the legal assessment. The distinction may be less glamorous than an AI pitch. For the person whose right to work depends on the filing, it is considerably more useful.

A case is a sequence, not a stack

Alma serves founders, researchers, engineers, other skilled professionals, and the companies employing them. Its menu covers temporary work visas including H-1B, O-1, L-1, TN, E-2, and E-3, as well as employment-based green card paths including EB-1, EB-2 NIW, and PERM-based filings. For a solo applicant, the service starts with choosing a viable path. For an HR team, the task grows into a portfolio: which employee is waiting on a signature, whose status expires next, what has been filed, and what will it cost?

The platform gives employers and employees a place to see those steps, share documents, and track action items. Alma describes a roughly two-week preparation timeline after evidence is collected. That last clause matters. Applicants still have to produce records, recommendation letters, and employer information; USCIS controls what happens after filing. A preparation target is not a promise that a visa will be approved in fourteen days.

The sales pitch is more concrete when told through customers. RoboForce, a robotics company, says its former law firm often took two or three days to respond, while case status and spending were hard to see. In Alma’s August 2026 case study, RoboForce says its people team now handles its H-1B, O-1A, and PERM program in roughly two hours a week, with a dedicated attorney responding the same day. Rundoo, a software company, reports spending less than an hour a week checking its immigration program even during busy periods. These are company-published customer accounts, not a controlled comparison. They do show the practical unit of value: fewer follow-up messages and fewer mysteries.

The price is part of the interface

Alma publishes a case fee schedule, which is refreshing in a field where the first bill can arrive with more suspense than the petition. A new O-1 matter is listed at $8,000. An H-1B cap or cap-exempt filing is $3,500. EB-1A and EB-2 NIW matters are each $10,000. Its fee includes attorney and paralegal support, platform access, compliance tracking, and employee communication. Government filing fees and outside charges, such as translations or evaluations, are separate. Larger employers can arrange custom monthly pricing; Alma also offers a split payment option to businesses.

Published Alma fees checked September 2026. Government and third-party charges are separate.

Flat pricing does two jobs. It helps a person judge whether to start. It also lets a company model the cost of keeping a team legally employed. Alma’s business product adds dashboards for case spend, milestones, and compliance deadlines. A law firm can be excellent at legal argument and still leave the buyer assembling its own spreadsheet. Alma wants to own both tasks.

The company learned to sell trust first

The pivot required money and distribution. In 2024, Marat wrote that Alma had raised a $5.5 million seed round in three weeks, after an earlier $500,000 pre-seed. TechCrunch had described $5.1 million in combined seed and pre-seed funding earlier that year; the figures come from different reporting moments. The investor list includes Bling Capital, Forerunner, Village Global, NFX, Conviction, MVP Ventures, NEA, and Silkroad Innovation Hub. That financing helped Alma build the larger operating model.

One detail in Marat’s fundraising account deserves a smile. The firm that provided the first seed term sheet had previously declined the pre-seed pitch, worrying about a missing technical co-founder and a small market. Alma returned with Chen and a bigger proposition. This was not a trick of wording. Selling software to immigration lawyers and delivering a complete service to immigration clients have different market sizes, costs, and obligations.

The first ten customers came from the founders’ networks and warm introductions. By February 2025, Marat said Alma had reached more than 150 customers. In between, the company tried social content, manual outreach, community events, and search traffic. Some outbound campaigns ran out of high-intent prospects; search took time; managing leads in Notion became unwieldy, so Alma adopted a CRM. There is a replicable lesson here for any professional service: referrals establish the first proof, but growth requires a system that remembers who asked for help and when.

This model has limits. A dashboard cannot create missing evidence. An attorney cannot make a weak petition strong by moving it faster. And an applicant’s case can still be delayed by government processing outside Alma’s control. The company reports a historical approval rate above 98 percent, but it also says prior results do not guarantee future outcomes. The fair claim is narrower and more interesting: Alma can make the work leading up to a decision easier to understand.

Alma sits between a traditional immigration practice and a software-only case tracker. It sells the legal service and the machinery that keeps that service legible. The company’s most revealing feature may be neither AI nor its dashboard, but the decision to let the customer see the bill and the state of the case before asking, again, for an update. In immigration, certainty is scarce. A clear next step is still a valuable thing.