1 tiếng nói · chưa gắn nhãn trưởng lab · 1 nền tảng · chưa có số tương tác · Nguồn crawl
I
Ilan Geller 1 bài viết · AWS Machine Learning Blog Nguyên văn nguồn Chi tiết nguồn
<p>As enterprises scale AI on <a href="https://aws.amazon.com/bedrock/" target="_blank" rel="noopener">Amazon Bedrock</a>, token spend often grows even as per-token prices fall. Without cost allocation tags in place, your cloud invoice shows AI spend as a single line item rather than by team, workload, agent, or model. Cheaper tokens expand adoption, and agentic workloads multiply invocations. As successful AI use cases scale, consumption can rise faster than unit prices drop. The economic question is whether the business value generated grows faster than the cost of the intelligence required to produce it. The Linux Foundation has launched the <a href="https://www.linuxfoundation.org/press/linux-foundation-launches-the-tokenomics-foundation-to-define-the-economics-and-roi-of-ai-value" target="_blank" rel="noopener">Tokenomics Foundation</a> to develop open standards for AI cost economics, a sign that managing the economics of AI is now an industry discipline.</p> <p>In this post, you will learn about six levers for optimizing token spend, how Accenture applied them in production, and how to apply them using Amazon Bedrock.</p> <p>Accenture applied these levers in four client engagements across telecom, insurance, mining, and a large enterprise deployment. The same levers are available natively on Amazon Bedrock as described in this post. Where savings were measured, token spend fell by about 40–85 percent. Results are specific to each client’s workload and may vary. Client names are withheld by Accenture to protect client confidentiality.</p> <h2 id="attribution-first-measuring-cost-at-the-business-level">Attribution first: Measuring cost at the business level</h2> <p>Before you optimize anything, you need to see where consumption is occurring, who owns it, and what business value it produces. Cost allocation tags at the team, workload, and agent level are the prerequisite for connecting consumption to accountable owners and, ultimately, to business outcomes. You apply them through <a href="https://docs.aws.amazon.com/bedrock/latest/userguide/cost-mgmt-application-inference-profiles.html" target="_blank" rel="noopener">Amazon Bedrock application inference profiles</a>, one profile per application or workload, tagged by team, cost center, and environment. After you activate them as cost allocation tags in the <a href="https://aws.amazon.com/aws-cost-management/billing-and-cost-management-console-home/" target="_blank" rel="noopener">AWS Billing and Cost Management console</a>, you can use these tags to trace which teams, applications, and agents are driving spend. They also make it possible to measure the impact of each change you introduce.</p> <p>Three AWS services provide the underlying telemetry. <a href="https://aws.amazon.com/aws-cost-management/aws-cost-and-usage-reporting/" target="_blank" rel="noopener">AWS Cost and Usage Reports</a> (CUR 2.0), delivered through <a href="https://docs.aws.amazon.com/cur/latest/userguide/what-is-data-exports.html" target="_blank" rel="noopener">AWS Data Exports</a>, handle financial attribution, breaking down billed inference cost by workload and team after your cost allocation tags are active. <a href="https://aws.amazon.com/cloudwatch/" target="_blank" rel="noopener">Amazon CloudWatch</a> captures operational signals such as invocation counts and latency. <a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-invocation-logging.html" target="_blank" rel="noopener">Amazon Bedrock model invocation logging</a> records request-level detail, including the model used and the input and output token counts for each call.</p> <p>Telemetry alone isn’t enough. Engineers can read cost per invocation on a dashboard, but product owners and program leaders need a metric expressed in business terms. Accenture’s Cost per Business Transaction (CpBT) framework bridges this gap. CpBT aggregates all the model invocations, retries, tool calls, and agent handoffs that contribute to one successful business output into a single fully loaded cost figure. It surfaces the true cost of producing a result, including the failed attempts that raw token pricing hides. Tracked against the business value each workflow delivers, CpBT distinguishes a growing bill that reflects scaling success from one that reflects waste.</p> <h2 id="the-six-optimization-levers">The six optimization levers</h2> <p>The goal of these six levers isn’t to minimize tokens at all costs. It’s to use the right amount and level of intelligence to deliver each business outcome at the appropriate cost, quality, and risk. The levers are organized into three phases based on the complexity of the change and the investment it requires. AWS and Accenture recommend working through the phases in order, and you can apply each phase independently to the workloads where it will have the most impact.</p> <h3 id="phase-1-configuration-only-no-infrastructure-change">Phase 1: Configuration only, no infrastructure change</h3> <p>With <a href="https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html" target="_blank" rel="noopener">Amazon Bedrock prompt caching</a>, the model reuses the parts of a prompt that repeat across requests, such as system instructions, tool definitions, and reference documents, so you stop paying full price to process the same content again. Amazon Bedrock stores the key-value state of previously computed prompt prefixes and is designed to skip recomputation on later requests. This is designed to reduce cached input token costs by up to 90 percent for supported models (see <a href="https://aws.amazon.com/bedrock/prompt-caching/" target="_blank" rel="noopener">Amazon Bedrock Prompt Caching</a>).</p> <p><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/batch-inference.html" target="_blank" rel="noopener">Amazon Bedrock batch inference</a> processes non-urgent workloads asynchronously as a group at 50 percent lower cost than on-demand pricing for select models (s