Three people, a handful of writing tools
Marianna and Dmytro Olefyrenko launched the first version of Easy-Peasy.AI in September 2022. The team was three people, the product was a handful of AI writing tools, and the goal — in the founders' own words — was simple: AI tools anyone can use, without code or complicated prompts.
It was a deliberately narrow start. In late 2022, AI tools were exploding in number but still mostly built for people who already knew how to prompt, how to chain models, how to read an API error. The Olefyrenkos bet the other way: that the larger audience was people who wanted the output — a blog post, a product photo, a voiceover — without learning the machinery underneath. Four years later, the company says more than 3.4 million people in over 150 countries have signed up, across 100+ languages.
The platform today is far wider than those first writing tools. It brings together MARKy Agent, AI agents for businesses, and the latest image, video and voice models in one place, with 300+ tools and templates. But the shape of the bet has not changed: one platform, every task, no code required.
The first interview, then a real company
The first outside validation came quickly. In February 2023, Japan's tech magazine ASCII.jp interviewed Dmytro about how Easy-Peasy.AI works — the company's first press interview, five months after launch. Six months later, in August 2023, the founders registered the company properly: Easy-Peasy.AI Pte. Ltd. in Singapore, as people joined from around the world.
That sequence matters. The interview came before the incorporation, and the incorporation came because the user base was already spreading beyond the founders' home market. The company was built in Singapore by a Ukrainian-born couple — Marianna runs the company and the product, Dmytro leads engineering — and it grew the way many Singapore startups do: international from the start, because the home market is small and the internet is not.
The two founders split the work along lines they had already spent a decade and a half building. Marianna brought 15+ years in IT across AI, software engineering, data analytics and UX design. Dmytro brought 16+ years in software engineering, including a stint as CTO of Kickoff by Zuju and leading frontend teams at BeMyGuest in Singapore. He built the first version himself, and still leads engineering — from MARKy Agent to the AI agents businesses use with their customers.
Awards, Google, and a trust score
The awards started in 2024. In July, Easy-Peasy.AI won Best Web 3.0/AI Startup in Singapore at the ASEAN Startup Awards. In September it was a finalist at the Global Startup Awards' Grand Finale in Istanbul. The same year it was named one of Singapore's top 500 SMEs in the Singapore SME 500 Award.
2025 added two more lines to the timeline. In July, the company was named a Brands For Good Honouree in the Technology for Good category. In October, it graduated from the Google for Startups Accelerator: AI First in Singapore — Google's accelerator for AI-first startups, and the kind of stamp that matters more to enterprise buyers than any award.
Then came the number the founders now lead with. In August 2026, Easy-Peasy.AI scored 84/100 in the Cybernews AI Trustworthiness Ranking — the joint highest score of 92 AI companies based in Asia, against an average of 60 for Asian companies and 65 worldwide. For a platform that asks users to hand over their brand voice, their customer conversations and their content, a trust score is not a vanity metric; it is the product. A month earlier, in July 2026, the company was named a Brands For Good Honouree for the second year running.
Founder-led, still teaching
What stands out about Easy-Peasy.AI is how founder-led it has stayed. Marianna is a certified trainer who teaches business owners and teams how to put AI to work — she has taught generative-AI bootcamps for Vertical Institute, spoken at the NTUC Career Festival 2026, and been a Google Developer Group member since 2019. Dmytro's open-source Vue libraries get 1.8M+ npm downloads a year, and he spoke at Google I/O Extended in Singapore in 2019.
That is the distribution strategy hiding in plain sight: the founders teach the market they sell to. Every workshop, bootcamp and open-source library is a chance for a potential user to meet the people behind the product before they ever see a pricing page. It is slow, unscalable-looking work — and it compounds. A founder who teaches a business owner to use AI this month is the obvious vendor when that business owner wants an AI employee next month.
The new AI Employees launch, live on Product Hunt in November 2026, is the logical next step: AI employees that work as one team and ask before anything goes out, on top of a platform that already does images, video, voice and agents. The founders' timeline suggests what comes next will look like what came before — a small, useful thing, shipped early, taught in public, and measured by whether strangers keep using it. From three people and a handful of writing tools, that approach has already carried them to 3.4 million users.
The playbook
- Start with the simplest useful thing — The first version was a handful of AI writing tools, built by three people. No sprawling suite on day one — one useful thing, done well, for people who had never touched AI.
- Teach while you build — Marianna teaches generative-AI bootcamps and business workshops; Dmytro maintains open-source Vue libraries with 1.8M+ npm downloads a year. Teaching is distribution and product research at the same time.
- Earn trust in public, repeatedly — ASEAN Startup Awards, Global Startup Awards finalist, Google for Startups Accelerator, and a Cybernews trust score of 84/100 — each one a public, third-party check on the product, collected year after year.