Real Estate & Facilities AnalyticsData Analytics Solution for Real Estate & Facilities Management Company in the US.
Problem Statement
The client is a US based NASDAQ listed company who owns, maintains and rents out over 10,000 properties to customers who are prospective residents. The company uses applications like Yardi, Salesforce and other custom applications for operations.
With complex leasing and maintenance process their reliance on technology makes them a market leader. They were using an on premises BI solution for their reporting & analytical needs. However, due to the rigidness of the traditional BI suite, it didn’t allow them to perform ad-hoc analytics at a rapid pace. They needed something where finance, operations, marketing, FPQA and other teams could mix and merge data themselves for quick analysis.
Challenges
As mentioned they use an on-prem conventional MSBI suite for their analytics which challenges them on the following fronts:
- Huge amount of maintenance and administration of Windows, File & SQL Servers leads to a bulky admin team adding to huge costs and overheads.
- Expensive licensing costs due to multiple on-premise servers involved with copies for HA & DR.
- Since SQL Servers and relational databases were involved scalability in processing wasn’t that easy.
- There were concurrency issues when there were parallel processes accessing the same data for reading and writing.
- Making changes in the data warehouse had several challenges since it was built on the ETL methodology rather then the ELT- Data Lake methodology.
- Maintaining multi-dimensional model had a steep learning curve and required highly-skilled resources to manage the infrastructure and take care of change requests.
- To make use of multi-dimensional models, clients had to learn MDX to analyze data in SSAS, which made data analysis tricky.
Solution Design & Implementation
We met the client and had a look at the current architecture that they had and also the in-house technical leadership as they wanted the solution to be migrated on a cloud service based architecture on AWS & Snowflake technology stack and using the data lake approach. Following were the salient features of the architecture that we designed and implemented for them:
1. Data Lake Architecture : A central data lake that serves as a repository of all data in its native from given by users, vendors, clients and partners.
2. ELT & Data Processing : The Extract Load Transform architecture allows the data in data to be transformed in its relevant form to be used by teams as deemed useful as per their use case.
3. Cloud Based Easy Access to Business Users : With the solution built on cloud, it helps users to easily query data from snowflake and easily upload files to be processed in AWS S3 which are browser based; this is much easier than the earlier server based on-prem technologies where business users required more training to perform their tasks.
0
Reduction in
Processing Time
0
Savings on
Human Capital
0
Up-Time
(Availability)
0
Infrastructure
Costs Saved
Future Plans
More Migrations
All the other on-prem reporting endeavors are now being migrated to the data-lake based cloud analytics framework.
Central Analytics
All new reporting comes through data lake allowing a central data repository helping business units do cross platform analytics.
AI - ML Initiatives
The data lake now allows the users to think beyond reporting and run AI & ML models to identify risks and potential opportunities.
Retire Old Servers
Due to migration, the old unused servers are not being decommissioned to save infrastructure costs and management efforts.