Nathan Leite Portfolio

About me

Playground

Schedule meeting

Home

Sofia

Sofia is a generative AI unified business hub that centralizes marketing, finance and many more tools for companies.

TIMELINE

2 years

Industry

Artificial Intelligence

OVERVIEW

Small entrepreneurs often manage marketing, finance, legal, HR, and daily operations alone, without formal training. Existing tools are fragmented and require specialist knowledge. This leads to decision fatigue, inconsistent execution, and lack of planning.

The market demanded intuitive AI solutions that seamlessly integrate into business workflows, oﬀering clear value without steep learning curves.

Recognizing this gap, I identified the opportunity to reposition Sofia by modernizing the user experience and expanding its capabilities with a mobile-first approach.

It was one of those moments where you realize the problem wasn’t complexity, it was clarity.

SUCCESS CRITERIA

Success is measured by clear product traction:

Higher feature adoption across core workflows.
Growth in conversion from free to paid tiers.
Improvement in perceived usability and visual clarity.

Features expectations

I conducted structured interviews with 40 high-value clients, synthesizing their workflows, pain points, and feature expectations. There was strong demand for innovative, easy-to-learn features.

Complex features were often abandoned due to poor discoverability.

Mobile usage was increasing but lacked dedicated design support.

Features expectations

I conducted structured interviews with 40 high-value clients, synthesizing their workflows, pain points, and feature expectations. There was strong demand for innovative, easy-to-learn features.

Complex features were often abandoned due to poor discoverability.

Mobile usage was increasing but lacked dedicated design support.

AI best pratices

During the workshop, I led the visual direction, ensuring accessibility and compliance to broaden adoption among less tech-savvy users.

Visual clarity and consistent feedback would reduce error anxiety.

Stepwise flows would improve task success and retention.

AI best pratices

During the workshop, I led the visual direction, ensuring accessibility and compliance to broaden adoption among less tech-savvy users.

Visual clarity and consistent feedback would reduce error anxiety.

Stepwise flows would improve task success and retention.

Stakeholder path + user feedback = $$$

I facilitated cross-functional workshops to establish clear priorities aligned with business objectives.

The requirements focused on balancing swift execution with quality, ensuring the design would support the imminent investment round and future growth.

Even with a clear path, I emphasized to stakeholders that ongoing user feedback was essential for continuous improvement.

Stakeholder path + user feedback = $$$

I facilitated cross-functional workshops to establish clear priorities aligned with business objectives.

The requirements focused on balancing swift execution with quality, ensuring the design would support the imminent investment round and future growth.

Even with a clear path, I emphasized to stakeholders that ongoing user feedback was essential for continuous improvement.

I led iterative sketching and prototyping sessions, exploring various AI navigation models. Early concepts were tested via remote usability sessions and A/B experiments to validate assumptions.

We refined concepts to minimize cognitive load, introducing clear aﬀordances and consistent visual language aligned with AI best practices.

Progressive onboarding improved feature adoption rates.
Mobile-first prototypes ensured responsive usability across devices.

I led iterative sketching and prototyping sessions, exploring various AI navigation models. Early concepts were tested via remote usability sessions and A/B experiments to validate assumptions.

We refined concepts to minimize cognitive load, introducing clear aﬀordances and consistent visual language aligned with AI best practices.

Progressive onboarding improved feature adoption rates.
Mobile-first prototypes ensured responsive usability across devices.

Technical feasibility

I evaluated the existing platform architecture to ensure design decisions aligned with technical constraints.

Key feasibility assessments included load times, API response limits, and frontend rendering capabilities.

This informed design iterations that balanced visual richness with lightweight code, essential for a smooth mobile-first experience.

This strategy result in <30% average page load time post-optimization.

AI behaviour

I designed Sofia’s AI interactions to lower the cognitive load by providing suggestion menus and pre-made prompts.

This allows users to act without needing to “know how to prompt,” which is essential for entrepreneurs who are overwhelmed and unfamiliar with AI language.

Guidance is the difference between feeling supported and feeling judged. The interface guides rather than demands, reducing friction and increasing adoption.

Key features

Fast Campaign Builder (Marketing)

Users answer a few questions about their business goals and audience and SOF.IA produces a full campaign plan, social posts, scheduling, and cost estimates.

Smart Legal Document Generator

Drafts service contracts, internal policies, and agreements tailored to local context.

HR Onboarding and Role Setup

Generates job descriptions, onboarding plans, and checklists based on the team size and role type.

Consulting Chat

A conversational advisor that asks clarifying questions and provides step-by-step action plans (ex: inventory, pricing, operations).

Micro Learning Mode

Short, AI-generated tutorials on key business topics (finance basics, target audience, contracts).

Sector-Based Personalization

Language, templates and examples change based on business type (beauty, food, home services, etc.).

FIRST STEPS

I focused directly into understand the main users and how their insecurities should be addressed.

FIRST STEPS

I focused directly into understand the main users and how their insecurities should be addressed.

Success metrics

Only the fine-tuned

best works.

R$300,000

New investment secured (Post-launch period)

41%

Adoption growth

+83%

Subscription revenue

12% -> 32%

Bounce rate reduced

ADDITIONALY…

I learned that introducing AI is not about adding more capability, but about removing friction. Real value comes from helping users act confidently without needing to understand the underlying complexity.

The outcome was not just an improved interface, but a more intuitive product direction.

Thanks for reading!