How Swatilina Barik Is Rebuilding US Immigration Strategy with Artificial Intelligence

By Sanya Kapoor – May 29, 2026

The problem: a fragmented, opaque visa system

The United States immigration apparatus has long been a bottleneck for tech talent, researchers, and specialized workers. Employers often face:

• Complex filing requirements – each visa class (H‑1B, O‑1, L‑1, etc.) has its own set of forms, supporting documents, and timing rules.
• High legal fees – law firms charge $5,000‑$15,000 per petition, a cost that scales with the number of hires.
• Unpredictable outcomes – approval rates hover around 60 % for H‑1B petitions, and the reasoning behind denials is rarely transparent.
• Lengthy processing times – even with premium processing, the average wait can exceed 60 days, disrupting hiring pipelines.

For startups that need to move quickly, these friction points translate into missed product launches and lost market share. The existing software tools merely digitize paperwork; they do not address the underlying uncertainty.

The solution: VisaArchitect’s AI‑driven platform

Swatilina Barik, a former immigration attorney turned tech founder, launched VisaArchitect in early 2024. The platform combines three AI components:

1. Generative document synthesis – using a fine‑tuned GPT‑4 model, the system drafts I‑129, I‑140, and supporting evidence sections based on a structured questionnaire. Users receive a complete, USCIS‑compliant draft in under five minutes.
2. Risk‑scoring engine – a supervised learning model trained on 1.2 million historical USCIS decisions predicts the probability of approval for each petition. The engine flags weak points (e.g., insufficient evidence of specialized knowledge) and suggests targeted improvements.
3. Compliance monitoring – a rule‑based layer cross‑checks deadlines, fee schedules, and labor‑condition application (LCA) requirements, sending alerts before any filing window closes.

The workflow looks like this:

HR manager → fills a short questionnaire → VisaArchitect generates a draft → Legal reviewer (in‑house counsel or partner firm) reviews AI suggestions → system submits the final packet via the USCIS online portal.

Because the AI handles the bulk of drafting, legal reviewers spend on average 70 % less time per case. Early adopters report a 30 % reduction in overall filing cost and a 15 % increase in approval rates for comparable petitions.

Funding and traction

VisaArchitect announced a $42 million Series B round on March 12, 2026. The round was led by Andreessen Horowitz, with participation from Cooley LLP Ventures, Founders Fund, and a consortium of immigration law firms that will act as strategic partners. The capital will fund:

• Expansion of the risk‑scoring model to cover EB‑2 NIW, EB‑5, and J‑1 categories.
• Integration with major HRIS platforms (Workday, BambooHR) to pull employee data automatically.
• A pilot program with the U.S. Department of Labor to test real‑time LCA validation.

Since its seed round in 2024, VisaArchitect has signed over 120 corporate customers, ranging from Series A startups to Fortune 500 enterprises. The platform has processed more than 8,000 visa petitions, with a cumulative $3.2 million saved on legal fees.

Why this matters for the broader ecosystem

VisaArchitect’s approach illustrates a shift from manual, lawyer‑centric processes to data‑driven, AI‑augmented workflows. The implications are twofold:

• For employers, faster, cheaper visa processing improves talent acquisition velocity, especially in sectors where the talent pool is global.
• For the legal market, the platform creates a hybrid model where attorneys focus on high‑impact strategy rather than repetitive drafting, potentially reshaping billing structures.

The success of VisaArchitect also puts pressure on legacy immigration software vendors, many of which still rely on static form‑fillers. Competitors will need to incorporate predictive analytics or risk scoring to stay relevant.

Remaining challenges

• Regulatory risk – USCIS could modify filing requirements or introduce AI‑specific disclosures, which would require rapid model updates.
• Data privacy – handling personal immigration data demands strict compliance with GDPR, CCPA, and U.S. federal privacy statutes.
• Model bias – the risk‑scoring engine must be audited continuously to avoid reinforcing historical biases against certain nationalities or education backgrounds.

Barik acknowledges these hurdles. "We are building a transparent audit trail for every AI recommendation, and we work closely with immigration experts to validate the model’s outputs," she said in a recent interview.

Looking ahead

With the Series B funding, VisaArchitect aims to launch a self‑serve tier for small businesses by Q4 2026, while the enterprise tier will incorporate real‑time case tracking and custom lawyer dashboards. If the platform can maintain its current trajectory, it could become a de‑facto standard for U.S. employment‑visa processing within the next three years.

---

Further reading

• Official VisaArchitect website – https://visaarchitect.com
• GitHub repository for the open‑source risk‑scoring model – https://github.com/visaarchitect/risk‑engine
• Andreessen Horowitz blog post on the Series B – https://a16z.com/2026/03/12/visaarchitect-series-b

---

Image credit: