James Damico, PhD, professor at Indiana University, has spent years researching how people consume information online. He wanted to create a tool that would help people regain control over how they consume information, teaching them to think independently. That is how the idea for YouVet was born: a platform where users evaluate real social media content and compare their results with people from around the world.

James came to us with a clear product vision. He understood what was needed to get started and was looking for a team that could quickly immerse themselves in the context.

1. Validate product hypothesis and launch an MVP

Get the product into real hands and gather live feedback instead of spending months polishing details.

2. Aggregate content from various social networks

Users needed to view existing posts from Instagram, Facebook, YouTube, and Reddit without leaving the application.

3. Deliver a frictionless user journey

A simple and intuitive mechanism for end users: view a post, vote on whether you trust it.

4. Build an analytics module

Voting results are broken down by age, gender, and region. Users can compare their own assessment with the opinions of others.

5. Create a content management tool

An admin panel where the team can add new posts from social platforms, organize them into categories, manage sources, and track how users interact with the content.

This project caught our attention before we even discussed the technical details. The idea of helping people critically evaluate information on social media is a mission that truly resonates with us. There was also a technical challenge involved, integrating with several closed platforms and processing diverse content types. But the choice was ultimately the client's.

Here is what influenced his decision:

1. Agile mindset

We have extensive experience working with startups and early stage products. When the client outlined his timeline, we immediately broke down the plan into phases and showed how we would move toward the goal.

2. Hypothesis-driven development

James came with a strong idea but without a rigid vision for every feature. We demonstrated our willingness to dive into the business logic and ask the right questions. The client valued this level of engagement, and it became one of the decisive factors.

3. Full-cycle partnership

Time was limited, and it was important for the client to have everything handled in one place. Technology selection, architecture design, development, infrastructure setup, and launch. We could offer all of this with a single team.

1. Parsing social media platforms

Social media platforms vary in their approach to content access: while some provide open APIs, others offer limited functionality or restrict access entirely. We developed our own parsers to collect content directly from the pages. This was complicated, however, by the routine layout updates that are par for the course on social media platforms. After each update, the parser would need to be rewritten, which was simply not a sustainable approach for us.

We built on Apify and wrote a dedicated scraper for each platform. To avoid rewriting everything from scratch whenever the layout changed, we moved all the element-binding logic into separate modules. When Instagram or Facebook updated their markup, we only touched the relevant module.

2. Diverse content formats

YouTube provides videos with descriptions, durations, and view counts. Instagram offers images or carousels with captions. Reddit features text posts with comments and votes. We needed to consolidate all of this into a unified format that would work seamlessly within the app interface.

We created a data middleware layer. Each parser outputs content in its own format, but before anything reaches the app, it passes through a converter. The app receives a unified structure: title, description, media, and basic metrics.

3. Tight deadlines

We had three months to deliver an MVP. The client came with a clear vision, but there was no detailed specification, and wireframes were being developed by another team. To meet the deadline, we needed to quickly understand the product and effectively prioritize our work.

We had three months, so we picked tools that wouldn't waste time. We threw together the admin panel in Retool in just a few days instead of building an interface from scratch. React Native lets us launch on both iOS and Android from one codebase. Basically, if there was a solid service we could use, we used it. We only wrote custom code when we absolutely had to.

4. Testing the app without publishing to the App Store

The client wanted to gather feedback from real users before the official release. However, Apple does not allow apps to be installed on iPhones outside of the App Store. For the average user, this means an app is either in the Store or it does not exist.

We used TestFlight for internal build distribution. Students scanned a QR code, and a signed version of the app installed on their iPhones. It was essentially full-scale user testing before the public release. This was our first experience doing it this way, and it paid off nicely.

5. Real time voting analytics

One of the key technical features of the application was built-in analytics. As soon as a user submitted a vote, the system needed to display up-to-date voting results. This meant we could not simply store votes in a database and calculate them once an hour. The data had to be aggregated instantly after each vote.

For basic behavioural analytics, we integrated Google Analytics. But that wasn't enough for the voting feature. We built our own analytics layer that aggregates results and breaks them down by age, gender, and region. We visualised everything through custom charts and Google Maps.

6. Bot protection

Bots can generate hundreds of votes within a minute, rendering the statistics useless. The conventional solution is CAPTCHA. However, we had to abandon CAPTCHA to improve the user flow, which meant finding an alternative approach.

Instead of captcha, we introduced SMS verification. Registration runs through Auth0, and number confirmation goes through Twilio. The user enters their phone number, receives a code, types it into the app (the whole thing takes just a few seconds). Meanwhile, bots get filtered out since they don't have real numbers. User data is stored on AWS using standard security protocols.

Frontend: React Native, JavaScript

Backend: Node.js

Database and infrastructure: PostgreSQL, AWS

Third-party integrations: Auth0, Twilio, Apify, Retool, Google Analytics

James approached us with an idea bigger than a typical startup. Within three months, he had a product ready to present to investors and users.

• Idea validation with real users. A group of students tested the app before the public release. The client gathered initial data on how people evaluate content and what they pay attention to.

• A social mission in action. The product began fulfilling its purpose. Users evaluated posts, saw the opinions of others, and compared their own results with the broader picture.

• Team autonomy for YouVet. The client received the tools needed to operate independently. The admin panel allows for adding content, managing categories, and monitoring votes. Day to day product management no longer requires developer involvement.

• Readiness to scale. The product is live and the infrastructure is in place. The client can now move forward by seeking investment, expanding the audience, and adding new platforms.

• Time and resources saved. Instead of a year of development with a full team, the project took three months with a small crew. Retool replaced a custom admin panel, React Native replaced two separate apps, and ready made services replaced in-house solutions.