The Spark Blog

Following my initial thoughts on the GenAI revolution, I've had the opportunity to speak with nearly 20 CIOs from large enterprises.

It’s fascinating to see how they are tackling the GenAI race to achieve both cost reduction and productivity gains.

While different verticals may approach this through various SaaS applications and services, the common ground is clear: they are shifting significant development resources away from other activities to focus on enabling Enterprise GenAI access to corporate data.

A recurring trend is the deployment of private instances of OpenAI's ChatGPT (or similar models from AWS) on Azure.

These enterprises are assured by Microsoft and OpenAI that the public models have no access to their proprietary data, and that the private instances are self-contained—unable to "learn" or "digest" their information beyond their walls.

The challenge, however, lies in allowing these private LLMs to securely access corporate data.

Several enterprises have reallocated 50 to 80 or more developers solely to bridge their private applications into RAG (Retrieval-Augmented Generation), which connects the LLM to their data. It’s a complex task involving significant coding, intricate logic, and VectorDB integrations to make this data accessible for RAG and the LLM.

The results?

Two global telecom leaders have integrated their customer service solutions with LLMs, leading to a 40-60% reduction in manpower. They believe with further optimization, they can improve this even further. Additionally, they’ve seen improved customer service efficiency and have begun automating upsell opportunities using GenAI-powered chatbots.

In the banking sector, two major institutions have developed next-generation chatbots capable of accessing private data in real time to offer customers personalized financial insights, saving strategies, and goal-oriented advice. This not only resulted in similar manpower reductions but also increased online revenue.

These are just a few examples of how, when implemented effectively, GenAI is transforming industries.



On the horizon, several other verticals are also gearing up to launch new GenAI-based services and processes. The race is accelerating, and it’s only a matter of time before these transformative solutions become ubiquitous across industries.