Overview
Taascom is a pioneering company dedicated to facilitating business transformation by helping other companies transition from traditional transaction-based models to subscription-based revenue models. This approach, known as "as a service" enablement, is designed to modernize business operations and revenue streams.
With the advent of generative AI, Taascom recognized an opportunity to leverage these technologies to enhance data analysis and provide intelligent recommendations to its users. The goal was to incorporate generative AI's data analysis capabilities into their application, enabling end-users to directly access insights and analyses without the need for specialized data analysts. This initiative was particularly aimed at frontline managers, such as those operating in quick service restaurants or overseeing sales in retail establishments, who could greatly benefit from immediate and actionable insights derived from their operational data.
To achieve this, Taascom partnered with Lyzr, a company known for its low-code agent framework that offers enterprise SDKs designed to integrate generative AI capabilities into existing applications or to facilitate the development of new applications utilizing Large Language Models (LLMs). This collaboration aimed to empower Taascom's end-users with advanced analytical tools, transforming their operational efficiency and decision-making processes.
Here is Ranga Raj, explaining Taascom’s decision to use Lyzr SDKs.
“We at Taascom wanted to provide a solution that would help SMB brick and mortar stores in retail and f&b markets improve their conversion rates leading to significant jump in sales and profitability. Our target customers were not too tech savvy and interpreting dashboards and taking operational and strategic actions was a tough ask.
To make this more consumable by store managers and owners we felt that generative AI was potentially option. But navigating through numerous options in this nascent and fast moving technology space was a tough ask.
Heard Siva in a webinar talking about how and why he started Lyrzr and it was similar to our need. Post a couple discussions we did a PoC on real data and we were pleasantly surprised at its capabilities to interpret the data and provide us the results we wanted. This coupled with GoML an implementation services company provided us the right vehicle to traverse through this fast changing complex environment to get a product out fast that would be capable of providing recommendations and actions that had significant business impact.
Would strongly recommend anyone embarking on a similar journey to reach out to Lyzr and hear how they can potentially help you.”
The Solution & Outcome
The introduction of generative AI into Taascom's operations brought with it a clear vision for enhancing user experience by enabling advanced data analysis capabilities. This technology promised a level of interaction and insight previously unattainable, allowing users to derive immediate value from their data. However, the transition was not without its challenges.
The primary hurdle faced by Taascom, and subsequently by their technology partner Lyzr, was the complexity involved in analyzing structured data. Taascom's extensive datasets were primarily housed in Amazon Redshift. This setup posed a significant challenge: creating an effective model capable of analyzing a large amount of data in real-time, converting text queries into actionable insights.
Lyzr's data analyzer SDK offered a solution to this challenge through two distinct methodologies: a Pythonic approach, translating text to insights (or queries to insights), and a text-to-SQL approach, enabling the conversion of queries to insights. By implementing these methods, Taascom was able to develop a data analysis engine that could interact with the Amazon Redshift data warehouse in real-time. This engine was capable of processing user queries, generating analysis results, providing insights, and offering end recommendations directly in response to user inquiries.
This innovation marked a significant departure from traditional data analysis methods, where users had to navigate through multiple dashboards, charts, and graphs to glean understanding from their business data. Now, Taascom’s users could simply pose their questions and receive immediate, insightful answers, streamlining their decision-making process and enhancing their operational efficiency.
The Enterprise GenAI Tech Stack