Audit your Azure, with ChatGPT.
LLMs (Large Language Models) have revolutionized the way we work, pushing us to alter our tool writing methods, software utilization, and our perception of privacy.
At work, I leverage Azure for various tasks. At home, I likewise use Azure for my personal projects, and infrastructure. While cutting-edge technologies like LLMs and ChatGPT have opened doors to innovation, they have also been misused by criminals. With advancements in voice cloning, image and video generation, and impersonation methods, scams have become frighteningly sophisticated.
Consider the alarming reality that in the hands of an unskilled scammer, these tools are highly effective. Now, imagine what a truly technical Machiavellian adversary could accomplish. I feel that trust in what we see and hear has been eroded so much faster in the last few years, we're sort of left to navigate this brave new world of uncertainty without wind in our sails or a compass to guide us.
Fortunately, there's hope. The very tools used to deceive can be harnessed to protect us. It's a massive shift, but we may soon see AI chaperones acting as digital guardians. In the elegant words of Deaner from Fubar, our AI would have the digital equivalent of "a giant laser gun, and it would sort of just be like, protecting..." And that's sort of cool, you know?
In the meantime, it's up to us to develop applications that leverage LLMs to enhance our efficiency. If you notice something boring, repetitive, and process-based in your daily work, it might be ripe for innovation and that's a good place to start looking to disrupt things.
领英推荐
My interest in MLOps is growing, driven by its exciting technological potential and lucrative opportunities in applications and automation. Recently, I explored a use-case for employing a LLM in security audits within Azure, creating a server that extends data retention past Microsoft's default periods. This project is vital for compliance, aligning with standards like NISPOM. Moreover, there is a cost component, as hosting the data locally is much cheaper than keeping it hosted in the cloud.
As a DevOps engineer, my exploration into Llama Index and SQL index capabilities has been fascinating. My audit project utilizes a Docker container interfacing with ChatGPT, enabling natural language queries on the database. Llama Index's SQL query generation from database schemas provides a handy tool for non-DBA professionals working on databases.
Embrace these new tools and adapt with them. The code I've created, crafted through collaboration with StackOverflow and ChatGPT, many google searches, and about thirty cans of Monster Energy, is shared in the hope that it aids anyone interested in joining this exciting change.