Litero — a next-generation AI assistant for writing.

Use the power of AI while keeping your unique voice and originality

Role

Lead UX/UI Designer

Platform

Web and mobile

Industry

EdTech

Duration

12 months

Litero.ai was created in response to the evolving landscape of education, accelerated by the rise of tools like ChatGPT. The goal was to offer students a comprehensive writing assistant that streamlines the entire process – from idea generation and sourcing literature to drafting, spell-checking, and plagiarism detection. Litero.ai was designed to support and accelerate student work without reducing it to mindless copy/paste approach.

Stage 1. Main business objectives

Quickly launch a fully functional first version (beyond MVP).
Analyze and optimize the conversion funnel to boost activation.
Find product–market fit fast and enable early monetization.
Rapidly test ideas and UI solutions to validate what works.

Stage 2. Research & Metric Definition

At the early stage, we used competitor data and analysis to define our market entry strategy:

Used publicly available metrics (conversion, engagement, retention) from a key competitor as benchmarks.
Conducted competitive analysis of similar tools.
Identified major weaknesses in existing products (like high volume of incoherent AI-generated text, use of outdated or irrelevant sources, poor plagiarism detection etc)

This gave us a clear baseline for product positioning and differentiation, as well as an understanding of key business metrics—such as customer acquisition cost (CAC), lifetime value (LTV), and the minimum lead-to-client conversion rate required to validate the project’s viability.

Stage 3. Design phase

After identifying our primary target user groups, we created a basic Journey Map to outline the most important initial touchpoints. We wanted to focus on the following key aspects of the experience:

Easy start: allowing users to begin working without registration, providing tips while building their workflow, and offering AI-powered suggestions automatically.
A touch of magic: creating an experience where the user makes minimal effort and immediately receives useful, digestible options.
Fast engagement: our hypothesis suggested that once users complete the initial onboarding and invest some effort into preparing their first project, it would be psychologically harder for them to abandon the process and close the site.

These were the main priorities that shaped the first product version.

Stage 1. Main business objectives

Quickly launch a fully functional first version (beyond MVP).
Analyze and optimize the conversion funnel to boost activation.
Find product–market fit fast and enable early monetization.
Rapidly test ideas and UI solutions to validate what works.

Stage 2. Research & Metric Definition

At the early stage, we used competitor data and analysis to define our market entry strategy:

Used publicly available metrics (conversion, engagement, retention) from a key competitor as benchmarks.
Conducted competitive analysis of similar tools.
Identified major weaknesses in existing products (like high volume of incoherent AI-generated text, use of outdated or irrelevant sources, poor plagiarism detection etc)

Stage 3. Design phase

After identifying our primary target user groups, we created a basic Journey Map to outline the most important initial touchpoints. We wanted to focus on the following key aspects of the experience:

Easy start: allowing users to begin working without registration, providing tips while building their workflow, and offering AI-powered suggestions automatically.
A touch of magic: creating an experience where the user makes minimal effort and immediately receives useful, digestible options.
Fast engagement: our hypothesis suggested that once users complete the initial onboarding and invest some effort into preparing their first project, it would be psychologically harder for them to abandon the process and close the site.

These were the main priorities that shaped the first product version.

Stage 4. Launch & Data Collection

Since polished UI was not a priority in early-stage development, we focused on speed over aesthetics. Used ready-made solutions (bootstrap component library, Material Design icons, no animations or advanced visual effects) helped us to reduce development time significatly.

Conducted 50+ user interviews with early adopters.
Analyzed over 100 session recordings via Hotjar to observe friction points and usage patterns.
Built a conversion funnel model and identified the drop-off stages with the highest user loss for further optimization.

Stage 5. Analysis & Iteration

The interviews and research provided critical insights and revealed key product weaknesses.
Improvements were prioritised based on their impact on product viability and user experience.
Weekly tracking of product metrics and behavioural trends shaped our ongoing roadmap.
Data-driven decisions guided both UX refinements and feature expansion efforts.

a black cellphone with a white letter on it

Conclusion

The project began showing signs of stable monetization around the 10th month of development. While the original product concept proved viable, the most impactful growth opportunities emerged later in the process. A combination of agile methodology and fast adaptation to new data allowed us to quickly validate ideas, kill non-performing concepts, and iteratively improve conversion at each step of the user journey.

Key metrics from the first 10 months:

~39% User Engagement Rate. This showed how often students relied on Litero.ai in their daily study routines, which proved the product was genuinely relevant to their academic needs.

~65% Document Completion Rate. Most users didn’t just play around — they actually completed full writing tasks, which confirmed the tool’s real-world usefulness.

<20% Drop-off Rate in Writing Flow. The low drop-off rate was a solid result for an early-stage product, showing that users found enough value to stay engaged through the entire process.

~5% Lead-to-Paid Conversion Rate. This turned out to be quite strong, especially compared to earlier competitors whose conversion rates hovered around 2.5%–3.5%.

B+ Academic Output Quality Proxy. Students consistently maintained good academic results — usually around a B+ or higher — which helped build trust in Litero AI as a serious educational tool.

Conclusion

Key metrics from the first 10 months:

~39% User Engagement Rate. This showed how often students relied on Litero.ai in their daily study routines, which proved the product was genuinely relevant to their academic needs.

~65% Document Completion Rate. Most users didn’t just play around — they actually completed full writing tasks, which confirmed the tool’s real-world usefulness.

<20% Drop-off Rate in Writing Flow. The low drop-off rate was a solid result for an early-stage product, showing that users found enough value to stay engaged through the entire process.

~5% Lead-to-Paid Conversion Rate. This turned out to be quite strong, especially compared to earlier competitors whose conversion rates hovered around 2.5%–3.5%.

B+ Academic Output Quality Proxy. Students consistently maintained good academic results — usually around a B+ or higher — which helped build trust in Litero AI as a serious educational tool.

Other projects

VeraViews - an end-to-end ecosystem for transparent advertising

VeraViews offers a comprehensive suite of solutions, including an Ad Exchange, SSP, Advanced Fraud Detection tools, and a Video Player, ensuring a secure and transparent ad ecosystem.

VeraVallet a crypto wallet for the VRA holders

A simple and secure wallet to earn, buy, store and stake VRA

Dimitri Wasilewski

Dimitri Wasilewski

Dimitri Wasilewski