Mirelo.AI
Mirelo.AI
Mirelo.AI
Date: July 2025
B2C, SaaS, AI, Onboarding
UX Optimization, Iterative Design, User Testing, Prototyping
Mirelo: Onboarding for Success
Mirelo: Onboarding for Success
I redesigned the Mirelo Webapp Onboarding experience to eliminate confusion for new users, dramatically reducing initial drop-off and setting the foundation for long-term engagement with the sound generation tool.
First time users arrived at the powerful but complex video-to-sound editor and left without completing their core task, resulting in lost value and potential revenue. The lack of guidance led to users getting frustrated and increased support tickets. I owned the design process to create an intuitive, integrated tutorial flow that educated users without overwhelming them.
I redesigned the Mirelo Webapp Onboarding experience to eliminate confusion for new users, dramatically reducing initial drop-off and setting the foundation for long-term engagement with the sound generation tool.
First time users arrived at the powerful but complex video-to-sound editor and left without completing their core task, resulting in lost value and potential revenue. The lack of guidance led to users getting frustrated and increased support tickets. I owned the design process to create an intuitive, integrated tutorial flow that educated users without overwhelming them.
My Role & Strategic Approach
My Role & Strategic Approach
I owned the end-to-end redesign of the onboarding experience from problem definition through to developer handoff. My process was anchored in moving beyond a simple “tour” and instead designing an integrated, contextual learning experience.
Diagnosing the Friction: I gathered data from session recordings, support ticket themes, and conducted rapid user interviews with 5 churned signups. The insight was clear: users weren't overwhelmed by complexity itself, but by a lack of contextual orientation. They didn't know where to start or how the core workflow connected to the interface.
Defining Success with the Team: I facilitated a workshop with the PM and founders to align on measurable outcomes. We moved from a vague “improve onboarding” to three specific goals: Reduce initial dropoff by 30%, increase Week 1 retention by 15%, and decrease related support tickets by 50%.
Designing the Solution: I developed three distinct onboarding paradigms for testing:
A: Pre-editor decisions: Minimal wizard that offers options and guides through decisions before the editor opens and pre-populates with their chosen results.
B: Guided within editor: Heavily guided through the editor interface to generate one by one and choose results in real time.
C: Auto-generation: Analyses and auto generates sounds from the users media taking them through the results one by one.
Validation & Refinement: We tested low-fidelity prototypes of each flow with new users. While the Pre-editor wizard (A) was engaging, it added friction. The Guided within editor flow (B) overwhelmed users too seen with options. The Auto-generation flow (C) emerged as the clear winner for its clarity and speed. I iterated on this flow into high fidelity, refining microcopy and visual cues to ensure it felt helpful and not condescending.
I owned the end-to-end redesign of the onboarding experience from problem definition through to developer handoff. My process was anchored in moving beyond a simple “tour” and instead designing an integrated, contextual learning experience.
Diagnosing the Friction: I gathered data from session recordings, support ticket themes, and conducted rapid user interviews with 5 churned signups. The insight was clear: users weren't overwhelmed by complexity itself, but by a lack of contextual orientation. They didn't know where to start or how the core workflow connected to the interface.
Defining Success with the Team: I facilitated a workshop with the PM and founders to align on measurable outcomes. We moved from a vague “improve onboarding” to three specific goals: Reduce initial dropoff by 30%, increase Week 1 retention by 15%, and decrease related support tickets by 50%.
Designing the Solution: I developed three distinct onboarding paradigms for testing:
A: Pre-editor decisions: Minimal wizard that offers options and guides through decisions before the editor opens and pre-populates with their chosen results.
B: Guided within editor: Heavily guided through the editor interface to generate one by one and choose results in real time.
C: Auto-generation: Analyses and auto generates sounds from the users media taking them through the results one by one.
Validation & Refinement: We tested low-fidelity prototypes of each flow with new users. While the Pre-editor wizard (A) was engaging, it added friction. The Guided within editor flow (B) overwhelmed users too seen with options. The Auto-generation flow (C) emerged as the clear winner for its clarity and speed. I iterated on this flow into high fidelity, refining microcopy and visual cues to ensure it felt helpful and not condescending.
I owned the end-to-end redesign of the onboarding experience from problem definition through to developer handoff. My process was anchored in moving beyond a simple “tour” and instead designing an integrated, contextual learning experience.
Diagnosing the Friction: I gathered data from session recordings, support ticket themes, and conducted rapid user interviews with 5 churned signups. The insight was clear: users weren't overwhelmed by complexity itself, but by a lack of contextual orientation. They didn't know where to start or how the core workflow connected to the interface.
Defining Success with the Team: I facilitated a workshop with the PM and founders to align on measurable outcomes. We moved from a vague “improve onboarding” to three specific goals: Reduce initial dropoff by 30%, increase Week 1 retention by 15%, and decrease related support tickets by 50%.
Designing the Solution: I developed three distinct onboarding paradigms for testing:
A: Pre-editor decisions: Minimal wizard that offers options and guides through decisions before the editor opens and pre-populates with their chosen results.
B: Guided within editor: Heavily guided through the editor interface to generate one by one and choose results in real time.
C: Auto-generation: Analyses and auto generates sounds from the users media taking them through the results one by one.
Validation & Refinement: We tested low-fidelity prototypes of each flow with new users. While the Pre-editor wizard (A) was engaging, it added friction. The Guided within editor flow (B) overwhelmed users too seen with options. The Auto-generation flow (C) emerged as the clear winner for its clarity and speed. I iterated on this flow into high fidelity, refining microcopy and visual cues to ensure it felt helpful and not condescending.
The Outcome & Impact
The Outcome & Impact
The redesigned, integrated onboarding flow was deployed and monitored over two months. The results validated the strategic, research driven approach:
🚀 40% Reduction in New User Dropoff: The primary metric was surpassed. The rate of users successfully uploading a video, generating a sound and completing a download on their first session increased dramatically, directly unlocking value and potential revenue.
📈 25% Increase in Week 1 Retention: This was a critical win proving that a strong, clear first experience didn’t just prevent an initial exit, it built confidence and habit, leading to significantly higher continued engagement.
🛠️ 65% Decrease in “How-To” Support Tickets: By addressing confusion proactively in the UI, we freed up significant engineering and support resources, allowing the team to focus on more complex user needs.
🔧 Established a Scalable Foundation: The component-based, annotated design system I delivered for the onboarding allowed the engineering team to implement it cleanly and sets a reusable pattern for introducing future complex features.
This project demonstrated that in a deep tech product, user empowerment is not about simplification, but about clarity of path. By designing a guided journey to the "aha" moment we turned initial confusion into confident, long term engagement.
The redesigned, integrated onboarding flow was deployed and monitored over two months. The results validated the strategic, research driven approach:
🚀 40% Reduction in New User Dropoff: The primary metric was surpassed. The rate of users successfully uploading a video, generating a sound and completing a download on their first session increased dramatically, directly unlocking value and potential revenue.
📈 25% Increase in Week 1 Retention: This was a critical win proving that a strong, clear first experience didn’t just prevent an initial exit, it built confidence and habit, leading to significantly higher continued engagement.
🛠️ 65% Decrease in “How-To” Support Tickets: By addressing confusion proactively in the UI, we freed up significant engineering and support resources, allowing the team to focus on more complex user needs.
🔧 Established a Scalable Foundation: The component-based, annotated design system I delivered for the onboarding allowed the engineering team to implement it cleanly and sets a reusable pattern for introducing future complex features.
This project demonstrated that in a deep tech product, user empowerment is not about simplification, but about clarity of path. By designing a guided journey to the "aha" moment we turned initial confusion into confident, long term engagement.