A Look Back at 6 Months of GenAI Implementation in Business

I'm Gary Rouch, co-founder and CTO of Qolaig, a company specializing in process automation with generative AI. For the past six months, we've been continuously engaging in client meetings and projects, and I'd like to offer a summary of what I've learned in this market, which is still in its infancy.

‍

The Rush for Use Cases

Today, it's rare to find a company that doesn't need to automate its processes, or at least optimize some of them. However, it can be challenging to identify which processes will bring real added value to the business. The companies that will stand out in this market are those that can identify the most relevant use cases.

At Qolaig, we quickly realized that most companies are not willing to invest significant sums in two-week audits, especially since these audits are not always necessary. Drawing on our experience and implementing BPA (Business Process Analysis) methods, we are now able, in most cases, to provide proposals after just three client meetings.

‍

A Technical Stack to Weather Any Storm

Once these processes are identified, the challenge is to implement them. For this, the technical stack plays a major role. It is generally composed of 3 components:

• An orchestrator such as Node-RED, Kubeflow, or Airflow for scheduling, data pipeline automation, and model retraining.
• Powerful and easy-to-use database tools, depending on the use case, such as Firebase or Elasticsearch.
• A cloud server such as AWS or Azure to host the solutions.

‍

Data Protection: A Major Challenge

Ultimately, data protection has become an absolute priority for most companies looking to adopt intelligent automation, due to the increased importance of regulatory compliance and privacy awareness. To address this challenge, various alternatives exist, such as using AI models that offer data protection guarantees, like Azure OpenAI, which provides a secure framework for processing sensitive data. And in cases where data security is paramount, some companies may opt for open-source models hosted directly on their own infrastructures, thereby ensuring total control over data management and security, while still benefiting from the advantages of automation and artificial intelligence.

Thank you for taking the time to read this article. I hope it has offered you enriching insights into the implementation of generative AI. If you're ready to explore how these innovative technologies can be integrated and leveraged within your company, or if you'd like to delve deeper into this conversation, please don't hesitate to contact me.