Save My AI Project

At Open InfoTech Solutions, we don’t consider ourselves an AI implementation agency. We see projects that use AI as workflow or automation projects first, and use AI strategically to achieve an outcome. If the goal is to use AI for everything, your project might not be the best fit for us. In short, we start with the goal and determine if AI is a tool useful for the task at hand.

New AI systems, specifically LLMs, are able to do amazing things right now that we wouldn’t have dreamed of even five years ago. They are enabling automation in many situations where it wasn’t cost-effective or even possible previously. That being said, a lot of AI projects are failing. An MIT study suggests that 95% of enterprise AI pilot projects are failing.

I think you’ll find a range of discussions and ways to interpret the data. Not here to promote our specific opinion, but there is plenty of evidence that the approach can drastically improve outcomes.
There are things that can be done to improve outcomes:

1. Use AI where it is necessary or provides value that exceeds any risk of false results.
2. Put guardrails around AI that don’t consist of AI. (Specifically, I mean workflows or automation that ensure inputs and outputs)
3. Provide visibility into the process so you can monitor outcomes and adjust.

At Open InfoTech Solutions, we’ve been building integrations and automations for years. With this experience, we now have a new tool in our arsenal: the LLM. Like all tools, it excels at some things and doesn’t work so well at others. Think of using your favorite crescent wrench as a hammer. It would sort of work, but it wouldn’t be the best solution to your problem.