Home Again MVP Web App

[jmulieri]

MVP Overview

The MVP aims to test the hypothesis that personalized service recommendations, driven by data analytics, will improve the effectiveness of homelessness service providers in securing permanent housing for participants. The core assumption is that individualized resource allocation will outperform generic approaches and ultimately lead to better housing outcomes.

Features & Scope

1. Interactive HMIS Dashboard

   • Users can explore and visualize local HMIS data, gaining insight into the homelessness situation in specific geographic areas.
   • Value: Helps service providers understand trends and resource availability in their region.

2. Unhoused Caseload Management (CRUD functionality)

   • Service providers can create, update, and manage participant records, including submitting enrollment information.
   • Value: Streamlines participant data management, saving time and reducing administrative burdens.

3. Personalized Recommendation Service Plan

   • Generate tailored service plans based on predictive models to promote permanent housing for each participant.
   • Value: Enhances decision-making by providing data-driven insights on which services to prioritize, maximizing housing outcomes.

4. Stretch Goal: Service Lifecycle Logging

   • Log services provided to participants over time, tracking progress and outcomes.
   • Value: Allows for longitudinal monitoring and data collection to further refine predictive models.

Key Questions Answered by the MVP

1. What is the state of homelessness in a given area, and what are the current and historical trends of PATH program participants?

   • Analyzes geographic homelessness data and provides tactical insights into the number of enrollments, services provided, and participant outcomes over time.

2. What services should I prioritize to maximize permanent housing outcomes?

   • Provides recommendations on which services have the highest likelihood of success based on individual participant data. Case managers will be able to view caseload and participant lifecycle data along with recommended service plan for increasing permanent housing outcomes.

3. What is the efficacy and performance gain of AI assisted service recommendation programs?

   • Offers insights into participant outcomes, comparing program success across traditional manually managed programs with those assisted by Home Again recommended services.

Solution Overview

The MVP for the web application will be hosted in AWS, making use of its services extensively. An overview of the components is defined as follows:

1. The backend API will be powered by a FastAPI service running within Docker on an EC2 instance.
2. A Postgres RDS instance will be used for the database with SQLAlchemy being utilized for the ORM (Object Relational Mapping) library.
3. The frontend will be implemented using React and the shadcn/ui component framework and hosted by a static website S3 bucket.
4. Recharts is a JavaScript charting library that is integrate into shadcn/ui and will be used for visualization on the application dashboard.