Product Overview


<aside> ❓ Getting your PVD Approved

</aside>

Persona(s)

Kim West is an aspiring social media influencer. She creates regular social media posts, with a blend of selfies, food pictures, and inspirational quotes. She eventually wants to move from a nano influencer status to become an influencer, with 100k+ followers, for the sweet sweet sponsorship money. But her time is limited, and her posts just could not get viral fast enough.

Problem

Kim West saw lots of dank content that has gone viral and wanted to make some of those dank memes. But she does not know how to start, what to write, and found it difficult to generate new dank meme ideas on a regular basis.

Mission

To help nano influencers go viral via generating dank memes and cool ideas for meme texts using artificial intelligence. Also, to help content creators to create and iterate the process of content generation.

Tech Stack Architecture

Memefly is a consumer app built on React native for frontend, using AWS cloud services and SQL/NoSQL database for backend.

For the data engineering stack, the data lake resides on AWS S3 with custom ETL packages that complete the data feed-back loop. The machine learning model is created with Tensorflow/Keras/Pytorch, packaged and deployed as a Restish API. The whole data and machine learning pipelines are then orchestrated with DAG tools such as Spark/Prefect/Airflow/Kubeflow.

Product Information


Vision - Problem & Objectives

Architecture Details

User Personas Detail

Solution & Features

Solution Details

Release Canvases

🔗 Link to deployed product

🔗 Link to staging product

🔗 Data Science Github

🔗 Front End Github

🔗 Back End Github

🔗 Trello