Sam.ai helps clients make decisions at the speed of thought with MySQL HeatWave

Enterprises have access to enormous amounts of data, but that data is often not easily accessible or curated for decision-makers. Currently, the workflow for business intelligence and reports resembles the functionality of a search engine: numerous results but no curation or context.

Sam.ai offers conversational intelligence and acts like an in-house data scientist that you can talk to and get answers 400% faster than clicks, taps, and keystrokes.  Sam HQ provides nuanced reporting that captures specificity and insights many levels deeper than existing engines—insights not limited by the search, sort, and filter options of a user interface. Sam is also able to answer questions from data from multiple sources through a proprietary process called Unified Holistic Data.

As can be expected, Sam.ai’s solutions require massive amounts of high performance computing (HPC) power to deliver accurate and meaningful results to user queries. From the start, Sam.ai used AWS RDS as its cloud and data management platform. Over time, the company began struggling with performance issues that often resulted in long response times or 404 errors for major clients with large datasets. For example, if Sam.ai could not answer a question in less than 30 seconds, it responded either that it couldn’t answer the question or that it didn’t have the data to answer it. Even after increasing computing spend by 35% and escalating issues via premier support channels, the company’s performance problem could not be resolved. Plus, management was keenly aware that any increases in computing costs would need to be passed on to customers. Consequently, Sam.ai evaluated alternative solutions to solve the challenges that threatened both company revenue and reputation.

Sam.ai’s leadership ruled out AWS based on previous experiences. Likewise, Snowflake was deemed to be too expensive. After a detailed evaluation, Oracle MySQL Heatwave Database Service (MySQL Heatwave) on Oracle Cloud Infrastructure (OCI), was selected on the basis of cost, performance, and support.

Oracle offered Sam.ai double its existing computing capacity at a 45% lower cost compared to AWS. MySQL HeatWave was the ideal solution for the company’s AI engine, especially with large enterprise client datasets. Additionally, Sam.ai’s leadership and technical teams gained premium support, including direct access to multiple levels of engineering resources.

After moving to MySQL HeatWave, Sam.ai doubled the processing speed it experienced with AWS at 45% of the cost. More than 128 GB of data were migrated to HeatWave and I/O operations per second  increased from 4,000 to 50,000 seconds. MySQL dramatically accelerated query speeds. For example, one query that previously took 41 seconds now is completed in 8-tenths of a second, representing a more than 50X improvement. In another example, a query that previously took 14.7 seconds now completes in 4-tenths of a second, a more than 35X improvement. Sam.ai’s production environment in OCI consists of 4 CPUs with 64GB memory and 1 TB of volume block for the database. For storage, the company allocates 12 CPUs and 1.5 TB, along with 1 CPU and 16 GB on Python.

With the radical increase in performance, Sam.ai now answers questions that it could not before. The combination of superior price-performance also gave the company a level of availability that it could not afford with AWS, Azure, or Google.

In the future, the company aims to answer the most difficult business questions, whether the answers generate millions of dollars of revenue or just make one user’s day better. Sam.ai is on a mission to have more and more clients come to know its offerings on a first-name basis.