Chai Point 用 Claude Code 做成本审计，月度 AWS 账单降 58%
Chai Point Cuts Monthly AWS Bill by 58%
印度连锁咖啡品牌 Chai Point 将月度 AWS 账单从 10877 美元降至 4550 美元，降幅 58.2%，其中约 80% 来自一名高级工程师用 Claude Code 执行的 Agentic Cost Audit。该审计以只读模式检查生产与 QA 云账户，定位闲置和孤立资源、过度配置及配置错误，并按节省金额、实施成本和风险排序。其余约 19% 来自更换 AWS 转售商 Isaga 获得 5% 折扣、以及 Savings Plan 降低约 15% 承诺支出，平台仍支撑 170 多家门店的 8000 多台联网冲泡设备。
正文
Beverage company and cafe chain Chai Point has cut its AWS cloud bill by 58.2%.
Dharmendra Raghuvanshi, Chai Point's director and head of digital transformation, shared the numbers in a LinkedIn post.
Raghuvanshi says the company’s monthly AWS bill fell from $10,877 to $4,550 over 12 months. “About 80% of it came from what we call the Agentic Cost Audit,” he says. The process was run by one senior engineer with Claude Code.
The audit used Claude Code to examine Chai Point’s production and quality assurance cloud accounts in read-only mode.
It matched individual cost items with usage data to identify idle infrastructure, orphaned resources, oversized deployments and configuration problems. It then ranked potential changes by estimated savings, implementation effort and risk.
The engineer implemented changes in stages, beginning with low-risk clean-ups before moving to configuration changes, infrastructure right-sizing and code-level optimisations. The team checked the subsequent AWS invoice against its estimates and addressed discrepancies in later rounds.
“AI did not replace our engineers’ judgement. It removed the cost of looking,” Raghuvanshi wrote.
The savings came without reducing the underlying platform's scale.
Chai Point’s Fountain technology stack, which supports its Intelligent Beverage Platform, continues to run more than 8,000 connected brewing bots across over 170 outlets, according to Raghuvanshi.
The company also reduced its cloud costs through commercial changes.
Between December 2025 and February 2026, it moved its AWS billing to Isaga, an AWS partner reseller, securing a flat 5% discount. Moving steady-state computing workloads to an AWS Savings Plan reduced committed spending by approximately 15%, the post said.
Chai Point attributed roughly 19% of the total savings to these commercial measures, with the remaining 80% coming from the engineering audit. The largest individual technical savings came from idle and orphaned infrastructure, which accounted for 16% of the reduction, and operational misconfigurations, which accounted for 15%.
Image: Dharmendra Raghuvanshi, LinkedIn.
The company now reviews cloud costs daily using an AI-assisted audit to identify unusual spending and further optimisation opportunities, according to a comment from Chai Point engineering manager Tanvir Alam on the post.