Hacker Newsnew | past | comments | ask | show | jobs | submit | techthumb's commentslogin

Amazon Bedrock Mantle provides OpenAI compatible API endpoints for model inference, powered by Mantle, a distributed inference engine for large-scale machine learning model serving. These endpoints allow you to use familiar OpenAI SDKs and tools with Amazon Bedrock models, enabling you to migrate existing applications with minimal code changes—simply update your base URL and API key.

https://docs.aws.amazon.com/bedrock/latest/userguide/bedrock...


Machine Learning capabilities analyze vast volumes of financial data, including transactions and check images to detect unusual customer behavior and suspicious activity with high accuracy and fewer false positives.


Transaction boundaries are a critical aspect of a system.

I've often noticed that these boundaries are not considered when carving out microservices.

Subsequently, workarounds are put in place that tend to be complicated as they attempt to implement two phase commits.


NASDAQ has moved the matching engine to AWS Outposts

https://www.nasdaq.com/press-release/nasdaq-completes-migrat...


I cannot help but think this is going to end up being a huge shitshow.

I hate AWS with a passion.


There's also https://jless.io


I use jq and jless and find that to be more practical than tools like these (not just because of being available in terminal).

When I have large blob of json, I am usually interested in first getting a high level idea of how it is structured (first few levels) then I either need to drill down or filter to find what I need. The combination of jless for interactive exploration and jq to make it easily scriptable/repeatable is quite nice.

I find it natural to reduce the json I start with to a subtree I can make sense of, and thus don't find myself lacking a tool which can provide a visual representation of the complete dataset.


K8S is not easy.

  It helps standardize:
    - deployments of containers
    - health checks
    - cron jobs
    - load balancing
What is the "old way" of doing things?

Is it same/similar across teams within and outside your organization.

If not, what would it cost to build consensus and shared understanding?

How would you build this consensus outside your organization?

For small organizations, one should do whatever makes them productive.

However, as soon as you need to standardize across teams and projects, you can either build your own standards and tooling or use something like K8S.

  Once you have K8S, the extensibility feature kicks in to address issues such as:
   - Encrypted comms between pods
   - Rotating short lived certificates
I don't love K8S.

However, if not K8S then, what alternative should we consider to build consensus and a shared understanding?


I love Cloudflare & their products.

From a moat perspective, I wonder if the TAM of this business in the context of the three cloud providers is substantial.

Their stock price is back down to reasonable levels.

Perhaps they'll remain a niche player in the market?


The TAM is definitely substantial, and IMO Cloudflare can become a top 4 cloud provider. They’re building a cloud network just like the big three, except they’re starting from the network layer and moving into storage and compute. The other clouds started from storage and compute, with networking either an afterthought or a price gouging mechanism that allows oversubscribing compute.

Cloudflare can move into the storage and compute market without changing their business model. Whereas AWS/GCP/Azure cannot suddenly reduce bandwidth prices or offer unmetered networking packages, because their business model depends on using bandwidth pricing to extract high margins.

From a practical perspective, as a developer building an app, Cloudflare has a lot to offer that the other clouds do not. Their new services are developer-friendly, built around modern ideals, and run on a global network. For certain workloads there is a clear argument in favor of choosing Cloudflare over one of the “big three.”

Yes, I see the stock is down. It reminds me of 2015 when AWS was still part of Amazon, and any developer could clearly see that the entire industry was using AWS. And yet Morgan Stanley published 30 page PDF analyses of Amazon’s business without a single mention of AWS. That’s when I learned that Wall Street investors don’t know wtf they’re talking about, which makes sense if you’ve ever met one. Cloudflare looks like a good buy to me.


    #!/bin/sh

    set -e

    if [ -n "$VERBOSE" ]; then
      set -x
    fi

    if [ "$1" = "--show" ] && [ $2 = "twitter" ]; then
      echo "https://twitter.com/teaxyz_"
    elif [ -n "$1" ]; then
      # Hi, I know you’re excited but genuinely, pls wait for release
      # I added this so I can do CI :/
      case $(uname) in
        Darwin) suffix=macOS-aarch64;;
        Linux)  suffix=linux-x86-64;;
        *)
          echo "unsupported OS or architecture" >&2
          exit 1;;
      esac

      if [ "$1" = "brew" ]; then
        d="$HOME"/.tea/bin
        mkdir -p "$d"
        curl https://tea.xyz/dist/tea-$suffix -o "$d"/tea
        echo "$d" >> $GITHUB_PATH
      else
        mkdir opt
        curl https://tea.xyz/dist/tea-$suffix -o ./opt/tea
        chmod u+x ./opt/tea
        shift
        ./opt/tea "$@"
      fi
    else
      echo
      echo "418 I’m a teapot"
      echo
      echo "thanks for your interest in tea."
      echo "alas, we’re not quite ready to serve you yet."
      echo
      echo "while you wait why not follow us on Twitter:"
      echo
      echo '    open $(sh <(curl tea.xyz) --show twitter)'
      echo
    fi


And a 404 for the macos build https://tea.xyz/dist/tea-macOS-aarch64

Same for the linux build https://tea.xyz/dist/tea-linux-x86-64


Good to see someone asking the important questions :)


https://status.aws.amazon.com hasn't been updated to reflect outage yest


There’s probably a lambda somewhere supposed to update it that is screaming into the darkness at the moment.

According to an internal message I saw, their monitoring stuff is fucked too.


Probably cached for good measure


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: