Procurement strategies for saving on your generative AI data project

Lou Salmen 13 Feb 2024 3 minute read

When procuring training and fine-tuning data services for generative AI, it’s important to understand their pricing models. Pricing can impact the quality of the data project stakeholders receive, which in turn affects the performance of their generative AI. But as a procurement manager, there are some creative ways that you can reduce costs without compromising data quality. 

Here are a few simple strategies you can use to maximize your organization’s data services spend.

Cross-leverage spends and services

Project stakeholders are often only familiar with the services their vendors provide to them – they may not be familiar with other services the same vendor provides to additional parts of your organization. For example, a pizza-buying department may have no idea that their vendor is also supplying their organization’s sandwich-buying department, and they could be missing out on significant benefits. 
 
That’s why it’s important for your procurement team to research and understand the full range of services that a data services provider may already be delivering to various teams across your organization. 
 
In many cases, one vendor servicing two departments is treated as two completely separate vendors by each stakeholder team within your company. That means benefits based on your company’s total spend with the vendor – such as lower service rates, volume discounts and favourable payment terms – are often lost. 
 
Make sure you negotiate a volume discount model with your AI data services provider based on your company’s total spend, rather than just your team’s spend. 

Propose volume discounts

Another thing to consider when it comes to vendors shared across teams is how your other stakeholders plan to use the AI data services provider, both now and in the future. 
 
Instead of relying on your provider to propose bands of spend for volume discounts, provide them with the volume discount tiers you’re looking for. For example, it’s not worth agreeing to volume discounts that start at $10M spend if you only plan to spend $2M with your provider. 
 
In general, the lower you set the starting point for volume discounts, the sooner you’ll realize opportunities to save and the earlier you can optimize your stakeholders’ budgeted spending plan with the provider.

Conduct pricing audits

Generative AI is evolving quickly, and new strategies for pricing and processes for training and fine-tuning them are routinely being introduced. 
 
You can optimize your spend by scheduling audits of your vendors’ pricing models and compare them against the initial projections they provided. Most AI data services providers make assumptions about the total cost of ownership of your generative AI upfront, but after completing several months of actual project work, they’ll be in a much better position to price more accurately. 
 
In some cases, you’ll be able to reduce your costs. In other cases, you might find that the pricing you negotiated at the start of the project explains the poor-quality data your provider delivered. 

Seek guidance from you AI data services provider

Instead of waiting for your AI data services provider to tell you about the latest and greatest generative AI training or fine-tuning strategies, ask them.
 
Schedule regular update calls to discuss creative ways to decrease the total cost of ownership of your generative AI. You might be surprised by the simple changes you can make on your end that could lead to major savings opportunities.
 
Navigating the procurement of generative AI data services isn’t easy. But a thorough understanding of pricing models, keen negotiation skills, regular pricing audits and open communication with your provider can all help optimize costs.
Embrace these strategies and you'll be well on your way to maximizing your organization's AI data services spend.
 
Not sure how to source the AI data you need to train and fine-tune your AI model? Download our AI data sourcing decision tree to get started.
Lou Salmen
Author

Lou Salmen

Strategy and Development Manager, TrainAI
Lou is Strategy and Development Manager of RWS’s TrainAI data services practice, which delivers complex, cutting-edge AI training data solutions to global clients operating across a broad range of industries.  He works closely with the TrainAI team and clients to ensure their AI projects exceed expectations.
 
Lou has more than 15 years’ experience working in sales and business development roles in the AI, translation, localization, IT, and advertising sectors. He holds a bachelor’s degree in Entrepreneurship/Entrepreneurial Studies from University of St. Thomas in St. Paul, Minnesota.
 
All from Lou Salmen