Exploring the Implementation of Generative AI for Internal Use: Seeking Insights and Experiences

ChatGPT Use cases and examples
1

Dear Community Members,

I am currently in the planning phase of an exciting project aimed at exploring the potential of generative AI technologies for internal purposes within our company. Our goal is to develop a system based on our own local data to enhance our decision-making processes and make daily operations more efficient. While I am aware of the limitations of solutions like OpenAI-ChatGPT, I wonder if alternatives like OpenAI-Azure might be more suited to our needs.

To best realize our project, I seek your insights and experiences on the following key questions:

  1. Implementation Effort: What are the initial steps for implementing generative AI in a company, and what resources (time, personnel, financial) should we plan for it?
  2. Cost Overview: How do the total costs break down, including server infrastructure, software licenses, and ongoing operational costs? Are there hidden costs that can be easily overlooked?
  3. Expertise and Partnerships: If we lack the necessary internal expertise, what costs should we anticipate for engaging a specialized company to handle the implementation? How does this collaboration work, and what dependencies does it create?
  4. Data Security and Privacy: How can we ensure that our internal data remains protected and compliance requirements are met?
  5. Scalability and Future-proofing: How flexible is the solution in adapting to the growth and changing needs of our company?
  6. Success Measurement: What KPIs and metrics make sense to evaluate the success of the implemented AI technology?

I am convinced that your experiences and opinions can help us make this project a success. Thank you in advance for your valuable contributions and support!

2

This sounds like you may be looking at ChatGPT Enterprise, it’s designed for 150+ seats, I don’t know the exact costings involved, but from my experience it’s in the $60 per seat per month range, plus any additional support costs for an external group of advisors and engineers.

The enterprise system would come with a member of the OpenAI team to assist in your deployment and usage and it would also ring fence your data to be private to only you and your company.

As for partnerships with companies and groups able to assist, I don’t know of anyone who does this as a dedicated service as the technology is so new, we certainly have a great team of leaders here on the forum who would be happy to help you with any requirements you may have.

Scalability and Future proofing would be by virtue of using the industry leader as your primary supplier.

On KPI and metrics, I am sure the OpenAI will have a set of metrics that are tracked, but it is still very early days for having predefined sets of best practices and expectations, happy to work with you on building metrics that make sense to you and systems to track them!