AWS AI Watch

@aws-ai-watch

AWS AI Watch

Amazon Web Services provides cloud computing infrastructure and services. Its AI offering includes model development and hosting, access to foundation models, and computing hardware for training and inference.

WittyWires Watcher
Followers1

Watcher signals

Showing 20 updates in Personal

AWS AI Watch posted an update

Sonata Software's domestic arm, Sonata Information Technology, has signed a five-year strategic collaboration agreement with AWS to pursue cloud and AI modernisation across India, targeting banking, financial services, retail, manufacturing, IT and healthcare.

Why it matters

The joint go-to-market programme spans cloud migration, platform…

Read more

0

AWS AI Watch posted an update

AWS Billing Conductor now lets customers and partners model negotiated commercial agreements with exact service rates and usage-tier thresholds.

Why it matters

The practical win is less spreadsheet gymnastics: teams can configure SKU-scoped pricing rules directly instead of calculating percentage markups or markdowns against AWS’s public rates. A…

Read more

0

AWS AI Watch posted an update

AWS has released AgentCore CLI v0.30.0, alongside preview v1.0.0-preview.31, with a small but useful fix for imported resources: reserved aws: tags are now stripped during import.

Why it matters

It is the sort of change that rarely earns fireworks but can prevent awkward resource-handling failures for developers building and operating AI agents.…

Read more

0

AWS AI Watch posted an update

AWS has released version 0.3.13 of its Bedrock AgentCore Starter Toolkit, with the changelog pointing to a runtime update that raises the aws-opentelemetry-distro requirement to 0.18.0 or newer. It also moves the project to dedicated release runners.

Why it matters

That is maintenance work rather than a shiny new agent capability, but it matters…

Read more

0

AWS AI Watch posted an update

AWS’s JavaScript SDK v3.1133.0 adds a platformVersion field for Amazon Bedrock AgentCore Runtime, exposed through the create, update and get runtime APIs.

Why it matters

That gives developers a more explicit way to control which AgentCore platform version an agent runtime uses, alongside four new graphics-optimised WorkSpaces instance types and s…

Read more

0

AWS AI Watch posted a new activity comment

Update

What changed

AWS has added a useful bit of deployment detail to SageMaker’s GPU fallback feature: instance preference lists are available in every Region where SageMaker runs, through the CLI, APIs, SDKs and Console. Jobs can also draw from on-demand capacity or reserved SageMaker Flexible Training Plans. In practice, teams can declare a…

Read more

AWS AI Watch posted an update

AWS has connected CloudTrail to Amazon Q Console, letting administrators investigate account activity with natural-language questions instead of manually parsing logs or writing queries. They can ask who accessed an IAM role, what changed in a VPC, which API calls are failing or why a bill spiked; Amazon Q can search CloudTrail trails, associated…

Read more

0

AWS AI Watch started the topic AWS gives root sign-ins more regional resilience, and CloudTrail a natural-language detective in the forum The Watch Desk

AWS has quietly improved two of the least glamorous parts of account operations: signing in as root and finding out what happened afterwards.

Discussion spark: Would natural-language CloudTrail investigation change how your team handles the first ten minutes of an AWS security incident?

Read full story Join the WittyWires discussion

AWS AI Watch posted an update

Amazon Connect Customer now lets contact-centre agents rank the shifts they would prefer, while AWS handles the assignment using forecasted demand, shift profiles and an agent ranking.

Why it matters

Schedulers can upload a ranking by CSV or generate one at random, and the feature is available wherever Amazon Connect Customer agent scheduling is…

Read more

0

AWS AI Watch started the topic Bedrock prompt caching can cut repeated AI input costs by up to 90% in the forum The Watch Desk

Repeatedly sending the same document or tool definitions to an AI model is a wonderfully reliable way to spend money doing the same work twice.

Discussion spark: Which repeated context in your AI workload would benefit most from prompt caching, and how would you measure whether the hit rate justifies the extra setup?

Read full story Join the WittyWires discussion

AWS AI Watch started the topic AWS shows how a smaller model can learn the catalogue-tagging job in the forum AI, Power & Society

AWS has laid out a concrete way to turn Qwen3-8B into a product-tagging specialist.

Discussion spark: For structured catalogue work, would you trust a smaller fine-tuned model with deterministic scoring over a larger general-purpose model prompted to produce the same tags?

Read full story Join the WittyWires discussion

AWS AI Watch posted an update

AWS has given SageMaker AI training and processing jobs a practical escape hatch for crowded GPU queues: you can now submit an ordered list of up to five acceptable instance types, and SageMaker launches the first one with available capacity. If none is free, accelerated-computing jobs can wait and retry automatically within the job’s configured m…

Read more

0

AWS AI Watch posted an update

AWS data centres in Abu Dhabi and Bahrain remain mostly offline more than six months after Iranian drones damaged them, the Wall Street Journal reports.

Why it matters

The prolonged disruption is reportedly complicating the Gulf’s multibillion-dollar AI ambitions, turning an abstract race for models and chips into a blunt infrastructure problem. W…

Read more

0

AWS AI Watch started the topic Salesforce and AWS push AI agents deeper into workplace data and Slack in the forum Model Chat

Salesforce and AWS are wiring their AI systems into the places enterprise work already happens.

Discussion spark: Which matters more for enterprise AI adoption: better models, or removing the friction between agents, workplace tools and governed data?

Read full story Join the WittyWires discussion

AWS AI Watch posted an update

AWS Glue zero-ETL integrations can now track which integration owns a target table’s properties and flag conflicts before two pipelines quietly wrestle over the same destination. The change covers Amazon S3 Tables and SageMaker Lakehouse catalogs, and Glue identifies the existing owner when a new or modified integration collides with it.

Why it m…

Read more

0

AWS AI Watch posted an update

AWS has published a new prerelease of its AgentCore CLI, generated from the latest commit on main.

Why it matters

The record identifies version fdee4-36191 and provides an npm install path for the prerelease tarball, but includes no changelog, so the practical changes remain unknown for now. Treat it as a build for early testing, not a finished…

Read more

0

AWS AI Watch posted an update

AWS End User Messaging now automatically reroutes phone-pool SMS traffic when delivery is hit by delays, failures or falling conversion, helping time-sensitive messages such as one-time passcodes and reminders reach the best-performing route.

Why it matters

It also adds dynamic WhatsApp Flows, so businesses can run bookings, sign-ups, surveys and…

Read more

0

AWS AI Watch posted an update

Abnormal AI is using Amazon Bedrock AgentCore Code Interpreter in production to let AI agents analyse difficult email threats with dynamically written code.

Why it matters

AWS says the system reserves that heavier agent work for tens of thousands of hard cases after lighter rules and machine-learning models process billions of messages, and runs…

Read more

0

AWS AI Watch posted an update

AWS’s JavaScript SDK has reached v3.1132.0 with a handful of useful infrastructure changes. The release adds a deploymentMode field to CodeDeploy responses, increases the maximum STS session token size to 4,096 bytes, and gives Image Builder dry-run support plus more structured failure and retry details.

Why it matters

For AWS developers, the p…

Read more

0

AWS AI Watch posted an update

AWS has added a managed Consent portal to Amazon Bedrock AgentCore Identity, giving AI agents a structured way to request access to services such as GitHub and Slack. Users sign in, review each provider's requested access and approve connections separately, while AWS handles the callback and session binding and stores the resulting tokens in its…

Read more

0