Hostinger Employee Auxe Zhirgule Discusses Where Founders Should Focus Instead of Coding

Artificial intelligence has significantly simplified the product creation process. For instance, Emilie Strimaitis noted that a shopping application, which took the development team almost a year to create in 2019, can now be built with an AI subscription over a weekend—and it will likely fail just as surely, but much faster.

Speaking at a webinar in partnership with Yourstory, which was part of VibeCode 30 at TechSparks 2026, the Head of Product Innovation at Hostinger pointed out that the initial steps founders took before the advent of AI remain relevant. The priority is still finding a problem, identifying the people who have this problem, and convincing them to pay. AI merely accelerates all subsequent stages.

Strimaitis emphasized that if someone can build a product over a weekend, thousands of others can do the same. He added that behind attracting attention has always been a real problem, even before the AI era. The danger lies in founders starting to build with AI, adding agents to run the product, and then agents to manage those agents, all without having a single customer. He compared this to a self-driving car with nowhere to go.

During the session, the audience participated in a poll where the most common issue when launching a digital product or online business was not knowing how to sell correctly, noted by 41% of participants. Strimaitis, who has independently launched about 15 ventures over 15 years, earning over $10 million without investors, discovered a constant pattern: 'Every success came from selling first, and every failure came from building first.'

He gave an example: a small tool sold before its creation cost about $100 to develop and brought in $29,000. Meanwhile, a platform built without demand confirmation cost over $40,000 and yielded no results. His rule is to use the same tools, the same AI, and the same automation, but in reverse order.

The first step is to choose a niche using three criteria: can you name the people, can you find them in one online place, and are they already paying for something? AI can help with this by turning your interests and skills into a list of potential ideas. Next, analyze complaints: you need to look for reviews and discussions in communities about poor experiences with existing tools. 'One-star reviews from other people are our business plan.' You need to gather about 50 such complaints, check if they are recurring, if the complainers are willing to pay, and if a better version can be created. AI tools can group these problems into core customer issues using their own phrasing. It is important not to let AI invent anything.

After that, formulate 10 promises or directions, such as 'faster,' 'cheaper,' or 'ready to use.' For each direction, create a simple landing page with a headline, description, and a pre-order or waitlist button. Then, choose one social platform and post content daily for 30 days, even if no one is watching. Record one long video weekly and cut it into parts. Those who register can be invited to a WhatsApp group, but only after they ask for it themselves. A group of 10–20 people becomes a research team and beta tester group. To attract an audience, you should genuinely interact with content creators who have approximately 5,000 to 50,000 followers before contacting them.

After 30 days, you must stop what isn't working and develop what is succeeding. If people have paid, fulfill the first 20 orders manually. Only after that should you start software development, and automation should be implemented last. Strimaitis openly stated that AI-based layoffs can turn against the author. Cold emails are losing effectiveness because everyone is now sending personalized messages using AI, so he recommends personally interacting with 20–30 content creators. On LinkedIn, he noted that posts marked as AI-generated rank lower. His alternative is to save posts that attracted attention and rework the idea in his own words.

For participants using Hostinger tools, the session mentioned using AI Builder for creating landing pages, Hostinger Agent for clarifying requests, and e-commerce for quick payment links. The main takeaway is simple: AI is a powerful tool, but demand must come first.

Giga+, formerly known as Alloha Fibra, confirmed that it has maintained conversations with Sky in recent months regarding a potential merger and acquisition. This information was disclosed in a statement to the market after B3 requested clarifications, prompted by a news item published by the Broadcast website, which belongs to Grupo Estado.

However, the company emphasized that there has been no deliberation by the company's management, nor the formalization of any agreement, commitment, or binding contractual document, as reported by the Teletime website.

A union of Giga+'s broadband operations with Zaaz, which is the fiber operator of Waiken ILW, the controller of Sky, would have the potential to establish the fourth largest fixed internet provider in Brazil, boasting approximately 1.91 million users.

In its statement, Giga+ declared that it has held dialogues with Sky and other industry players in recent months. The operator also mentioned that it conducts periodic evaluations of business opportunities and strategic alternatives aimed at expansion and capital management.

Such a transaction could be advantageous for Giga+, given that it registered a 16.5% drop in its access base during the twelve months ending in August. In contrast, Sky and Zaaz's broadband operation showed an expressive growth of 47.7% in the same period.

In addition to operating in the fiber optic network, Sky is seeking to penetrate the mobile network segment. In August, the company officially launched the commercial plans for Sky Móvel, which operates as a virtual operator (MVNO) using Surf Telecom's infrastructure. There are also plans to introduce Sky Link, a competitor to Starlink in the satellite internet market.

An eventual operation would include Giga+ and the broadband assets of the Waiken ILW group, which owns Sky and Zaaz. According to data from the consultancy Teleco, Giga+ had about 1.29 million accesses in August, while Sky and Zaaz totaled approximately 627 thousand customers. Combined, these operations would reach about 1.91 million subscribers.

This volume would position the consolidated group behind only the large companies Claro, Vivo, and Nio—the latter having absorbed Oi's fiber clientele—but ahead of Brisanet, which had about 1.58 million accesses at the same time.

The geographical reach would also be expanded. The two operations would operate in 1,110 combined cities, leading in 173 municipalities. Currently, there is a network overlap between the companies in only 123 cities.

Currently, Brisanet holds the fourth position, slightly ahead of Brasil TecPar, which has about 1.38 million accesses. The latter also aims to increase its market share through the acquisition of Ligga Telecom's operation, a deal that has already received approval from Anatel and the Administrative Council for Economic Defense (Cade), but which is still awaiting a ruling from the Ministry of Communications.