Valid Social Influencer Audit

An AI-powered dashboard to validate if a social media influencer authencity based on the followers, engagement, and other parameters.

Comment

Inspiration

Markering with influencer has become huge billion dollar industry. Some brands lose millions every years due to influences fraud, ghost accounts and inflated metrics. Real time verification for indluencer is challengind during crucial marketing decitions. This ValidSocial is objected to bridge the need for auditing engagement authenticity before deploying marketing budget.

What it does

It is B2B SaaS that designed to analyze and audit social media profiles of an influencer. Their authencity will be "Flagged" or "Verified" based on account activities. Users from bussinees marketing can input username, follower account and engament rate. The AI Scoring System will compute the AI Fraud Score and give the 'flag'.

How we built it

We engineered ValidSocial utilizing a modern full-stack serverless architecture:

  • Frontend Framework: Next.js 14+ (React), Tailwind CSS, Shadcn UI, and animated charts using Recharts.
  • Backend Infrastructure: Next.js Serverless API Route Handlers integrated with the official AWS API.
  • Database Layer: Amazon DynamoDB
  • Deployment & Hosting: It is deployed fully on Vercel with connection to dynamoDB.

Challenges we faced

  • API Constraints & Simulations: Public data from strict graph APIs (like Meta) within a short hackathon window presents high administrative barriers to accessing.

Accomplishments that we're proud of

  • Successfully deploying a live full-stack app integrated directly with Amazon DynamoDB. *Simple MVP for users to check the authenticity of influencers.
  • Implementing a dynamic backend AI logic engine from the input to expected ouput.

What we learned

Experience in connection of vercell, backend with NoSQL storage in AWS. Some challenge ini build the initial environment.

What's next for ValidSocial

  • Direct Socmed API integrations:
  • Advanced NLP Analytics via AWS: Integrating Amazon Comprehend to run machine learning sentiment and spam analysis on profile.
  • Automated PDF Reporting:

Built With

Share this project:

Updates