Homeless Services Dashboard

Goal

My homeless services dashboard provides a comprehensive view of system performance by tracking shelter usage, program enrollments, and client outcomes across key services. It transforms raw data into actionable insights through interactive visuals, enabling stakeholders to monitor trends, identify gaps, and improve service delivery.

Tools

AWS S3 | Snowflake | Power BI | DAX | SQL | M Language

Key Insights

  1. We’re seeing a consistent rise in new enrollments during the winter months, suggesting a need to scale services seasonally.
  2. Roughly 1 in 5 clients re-enter within 90 days — a signal that permanent housing support may be insufficient or follow-up services underutilized.
  3. Program B has the lowest average length of stay and lowest recidivism, indicating a potentially replicable service model.

Process

I developed a homeless services dashboard as a simulation project using mock data to demonstrate how data engineering and analytics can support more informed decision-making in the homelessness response system. The goal was to model how individuals interact with services like emergency shelters, coordinated entry, and housing programs, and to highlight where bottlenecks or service gaps may occur.

The process began with storing raw CSV data in an AWS S3 bucket, simulating a cloud-based intake system. From there, I used Snowflake to build the data warehouse. I designed and implemented the database schema to reflect key entities—clients, services, enrollments, and outcomes—and used SQL to clean, normalize, and join the datasets into a usable format. This included standardizing date fields, creating lookup tables, and ensuring referential integrity across records.

Once the data pipeline was in place, I connected Snowflake to Power BI to begin building the reporting layer. Within Power BI, I used Power Query and M language to perform additional transformations and shape the data for analysis—such as splitting complex columns, creating calculated tables, and filtering for active records. I then used DAX (Data Analysis Expressions) to calculate core metrics like length of stay, recidivism rates, enrollment flow, and program utilization.

The final dashboard features a series of interactive visuals that allow users to explore the data by filtering across time periods, demographics, service types, and outcome measures. It’s designed to be intuitive and actionable—making it easy for stakeholders to identify trends, monitor performance, and make data-informed decisions about where to focus resources.

Although the data is fictional, the dashboard simulates a realistic workflow for a homelessness data reporting environment. The project demonstrates a full end-to-end data pipeline—from raw data storage and warehousing to visualization and insight generation—using tools like AWS S3, Snowflake, SQL, Power BI, M, and DAX.