Ashvin Vellody says firms track per-workflow and per-query costs
As AI token prices climbing, companies are rethinking how they spend on tech.
Ashvin Vellody, chief strategy and innovation officer at Deloitte South Asia, shares that firms now track things like cost per workflow and customer query to make sure their investments actually deliver value.
To keep budgets in check, businesses use dashboards, set usage limits, and require approvals for pricey models.
Companies choose local models and GPUs
To save money, companies like HCLTech are choosing local AI models and GPUs over expensive external systems, says CTO Vijay Guntur.
Tech Mahindra is using smaller open-source models and building its own AI models for their needs, according to Chief Innovation Officer Nikhil Malhotra.
Since global token use is expected to skyrocket by 2030, frequent checks help prevent overspending and keep costs under control.