NoSQL Database | BART Delivery Analysis
Client: AGM Fresh Foods
Role: Data Science & Product Innovation Fellow
Duration: Summer 2024
Contributer(s): Averine Sanduku
DISCLAIMER: AGM is a fictional business in the Bay Area please note that this project was completed to showcase my skills NoSql database skills and analytics.
Overview
Deliver a AI-powered food delivery optimization using graph algorithms and AWS to enhance customer experience and reduce delivery latency.
Acme Gourmet Meals – Smart Delivery Optimization
Acme Gourmet Meals is an AI-enhanced food delivery platform designed to optimize last-mile logistics for a high-end meal service. The project leverages graph algorithms and cloud infrastructure to identify ideal pickup locations near BART stations, reducing delivery time and improving customer satisfaction. By combining real-time data, spatial analysis, and community detection techniques, the system enhances route efficiency and supports a scalable delivery model.
Corporate Culture & Values at Acme Gourmet Meals
A customer-focused, innovation-driven food service committed to excellence, sustainability, and team growth—delivering gourmet quality with health, wellness, and adaptability at its core.
AGM Business Case Scenarios
To expand Acme Gourmet Meals’ reach across the Bay Area, our team focuses on optimizing pickup location placement through BART integration, enhancing user accessibility via a mobile app, and preparing for scalable delivery with NoSQL database solutions. While robotic delivery is on the horizon, our current priorities lie in smart logistics, user experience, and adaptive data infrastructure.
Business Case Scenarios
PickUp Locations | Distribution Points | BART Integration
Why Use Neo4j?
We chose Neo4j, a leading graph database, to model and analyze the BART transportation network due to its ability to efficiently represent complex relationships between distribution points, BART stations, and delivery zones. Traditional relational databases would have made querying route connectivity and identifying optimal pickup locations significantly slower and more cumbersome. With Neo4j, we leveraged graph algorithms like Louvain Modularity and Betweenness Centrality to detect community clusters and high-traffic hubs—critical for strategically placing delivery pickup points and improving route efficiency. Its native support for nodes, edges, and fast traversals enabled us to build a scalable, data-driven logistics system tailored to urban infrastructure.
Leveraging Graph Algorithms for Strategic Decision-Making in AGM's BART-Based Network
Louvain Modularity is a technique used to identify communities within large networks by maximizing the modularity score. This score measures the strength of the division of a network into modules, where a high modularity indicates dense connections between nodes within the same community and sparse connections between nodes in different communities.
Using the Louvain algorithm, we identified 9 distinct communities within the BART network. These communities represent clusters of stations that are more densely connected internally than with the rest of the network.
Identifying these 9 communities allows us to effectively bucketize our network into high-level regions. This division simplifies our operational strategy by creating manageable segments, making it easier to cater to the specific needs of each region. As a result, we can ensure that our pickup points are optimally placed within these communities.
Business Case 1: Personalized Recommendations
What we used to achieve this?
MongoDB
What is Mongo DB?
NoSql Database known for its flexiblity, scalibility, and performance. Designed to handle large volumes of unstructured data and supports a variety of data types, making it a popular choice for modern applications that require rapid development agility.
Business Case 2: Enhanced Delivery Tracking
What we used to achieve this?
Redis
What is Redis?
An open-source, in-memory data structure store that is often used as a database, cache, and message broker. It is known for its speed, flexibility, and wide range of data structures, making it ideal for applications that require real-time performance and high throughput.
