What special considerations are needed when implementing Generative AI tools at the enterprise level?

588 views1 Upvote6 Comments
Sort By:
Oldest
Chief Strategy Officera year ago
Focus on solving for the use cases and business requirements instead of getting caught up in the hype. Do not be fooled by a great POC. There is a stark difference between how a Generative AI tool is perceived in the lab versus when it is finally rolled out to your employees and customers and integrated with existing workflows at an enterprise level. 

 

Business and technology leaders will need to collaborate and strategize on the best approach which is ‘human-centric’ : 

Focus on and address questions of responsible and ethical use of AI (e.g., data bias, proprietary information, etc.)
Decide whether you will use large publicly available models with some APIs and consideration towards cybersecurity or initiate smaller very niche models for specific use cases which are fine tuned to meet your requirements.
Include the team by educating and clarifying that this implementation will augment but not replace human effort. The idea is to empower the employees and serve the stakeholders better more effectively and efficiently.
2
Chief Strategy Officera year ago
Would like to know the views of other leaders in the community about the issues faced and mitigating actions undertaken in their respective field of work ? 
Associate Director of Data Science & Analytics in Healthcare and Biotecha year ago
Cost. Training these models is extremely expensive. A large internal LLM can cost tens of millions of dollars in compute. And that is assuming your organization has the proper infrastructure to support it, and the right expertise to optimize it. 
lock icon

Please join or sign in to view more content.

By joining the Peer Community, you'll get:

  • Peer Discussions and Polls
  • One-Minute Insights
  • Connect with like-minded individuals
CIO in Insurance (except health)7 months ago
For organizations operating in highly regulated environments (e.g. healthcare), governance and compliance is a critical consideration.

For all enterprises, training at all levels is critical. These tools require a fundamental shift in thinking especially critical thinking. Developing skills in prompt engineering, Gen-AI oversight, specifically how to detect when the LLM/tool is straying from the task or hallucinating.
VP - Chief Architect in Banking7 months ago
Every organization is a mix bag of vendor supplied technology and home-grown stuff. When it comes to vendor saturation (e.g. supplied platforms/tools/apps) we're soaking at around 90%-95%. The majority of our tech vendors are planning to implement some form of Generative AI in their next release. So, the consideration becomes... how to manage the saturation of AI capabilities across the vendor technology landscape. All require one critical thing to work.... data. It's more than policy controls or contract management. It's like we need AI to solve for the AI saturation. LOL

Content you might like

Cost of RPA products27%

Lack of developers who can code RPA applications44%

Amount of customization needed to automate business processes24%

Lack of RPA code maintenance resources4%

View Results
11.7k views5 Upvotes8 Comments

Data quality (e.g., limiting duplications)19%

Business processes and workflows (i.e. data flowing between systems)65%

Data accessibility (decreased wait times for access, access logs for auditing)12%

Regulatory compliance

Other4%

View Results
265 views