eva.ai

Valvis Brogas spent years in business development, sales, and project implementation roles, helping clients achieve over $100 million in cost savings. During this time, he noticed a persistent pattern across industries: customer support was bloated, expensive, and inefficient. Whether it was e-commerce brands dealing with product inquiries, travel companies handling booking amendments, or hospitality brands processing complaints, the underlying problem was identical—too many inquiries, not enough smart automation. That's when Valvis and his co-founder Zeno decided to build eva.ai, an AI-powered customer support solution that could dramatically cut costs while improving response times.

The team started from scratch in May/June 2024. "When we incorporated somewhere in that time. So basically four or five months, we built from kind of a scratch on the paper and having an idea to actually building a functional and repeated" solution, Valvis explained. The vision was bold: cut customer support costs by as much as 97% and deliver responses in 10 seconds or less using AI and automation. But building in the AI space in 2024 was noisy and saturated. Rather than rushing to launch, the team spent months narrowing down their focus. "In the beginning we wanted to narrow down what we're doing because right now the whole kind of when we talk about AI space, automation space, it's very saturated. There's a lot of noise right now." They eventually chose to target three initial niches: e-commerce, travel, and hospitality—industries they understood deeply from previous work.

The team grew to 6 people, bootstrapped entirely on personal savings. "They've each put in several thousand dollars to help fund the business," with Valvis contributing "a few thousand" himself while maintaining consulting work on the side to cover living expenses. His co-founder Zeno had additional income streams to support the effort.

By November 2024, eva.ai was about to land their first two paid pilots—a milestone five months in the making. The deals came through relationships and introductions, not cold outreach. "The pilots actually came from introductions and from clients that I worked and served earlier before that I worked in my previous jobs," Valvis said. One pilot was structured as a $5,000 flat fee upfront for full integration and implementation—a complete support automation setup with no need to hire new support staff. The second was offered at their lowest pricing tier: $149 a month, allowing them to test which industry and use case would be the strongest market fit.

When pressed on why progress took five months to reach the first customer, Valvis was honest: they spent considerable time debating positioning. The founder acknowledged Nathan's valid criticism that three niches (e-commerce, travel, hospitality) was still too broad, but argued they needed to test across segments quickly to find where they'd excel. "We know that we're going to be definitely excelling in one more than we do in the others. And I think that's what we want to quickly test it out and just go for that one."

eva.ai enters their paid pilot phase with two very different contract structures—one testing a high-touch, flat-fee enterprise model ($5,000), the other a lightweight, usage-based approach ($149/month). The team is using these pilots as learning labs to determine which industry segment and pricing model will drive the strongest unit economics. With 6 people, limited runway, and serious skin in the game (Valvis has family obligations, two kids, and a wife also working), the pressure is on to prove the AI customer support market wants what they're building. The next few months will tell whether they've identified a real market gap or merely built another feature in a crowded AI space.