Tim — your cloud on autopilot

Jason Kim
5 min readNov 22, 2018

Today, we’d like to introduce you to Tim — Taloflow’s Instance Manager. We created Tim to take the stress out of managing your AWS bill. Every year, the major cloud platforms have been making big usability enhancements and releasing features such as Serverless that make it easier and easier for developers to get started, spin up, and scale resources. Despite those efforts, getting clarity on your cloud bill and figuring out how to reduce and rightsize your cloud spend is still no easy feat today. To tackle this problem, we built Tim, an AI that will help monitor and manage your resources and make cost management as seamless and as “autopilot” as possible. With features such as auto-discovery, auto-tagging, tailored recommendations, Slack automation, and napping — Tim makes it easy to ensure your resources are being properly utilized.

The cloud bill puzzle

The cloud has brought many benefits to developers, startups, and enterprises alike, making innovation faster and achieving scale easier. However, the ripple effect of the cloud has created a big challenge for companies, as the entire paradigm of infrastructure budgets has changed. Instead of a CIO or CTO provisioning infrastructure (ie: 5 million dollars for X machines and Y data centers), infrastructure spend has become completely dynamic and variable. This forces organizations to treat the cloud as a utility, forcing teams into an entirely new model for cost optimization where the developers and engineers have the levers but lack the tooling to effectuate proper change. Getting clarity on how you’re spending on AWS can also be a complete mystery, let alone figuring out if you’re spending your budget in the right way.

Tweet from Nicholas C. Zakas ( Creator of @getslint)

To understand how teams are handling this problem today, we surveyed dozens of SRE, DevOps, and Engineering teams to see how this problem is being handled today.

Here were the major issues that were quoted:

  • Teams don’t understand their cost impact on infrastructure and enforcing tagging of resources for accountability is cumbersome”
  • Feedback cycle for spend is too slow (3–7 days for certain tools or monthly on Amazon) so we’re always reactive, and the cost metrics ignore hidden costs”
  • Current tools over complicate things and have a big learning curve, cost management is already complicated and if your tool makes things more complicated our developers just won’t use it”
  • “I’m building a kludge between Tableau, DataDog, and CloudHealth to tie in infrastructure costs to our own internal cost centers and processes”`

This informed our product-thesis and first principles in designing Tim:

  • We need to simplify cloud cost accounting and management and reduce the friction for teams to adopt best practices and tools.
  • We need to design with the developer in mind and as they are the ones with the lever impact cloud spend.
  • We need to provide real-time insight into their cloud spend.
  • We need to connect disparate systems and data sources to provide one unified interface to deal with infrastructure spending.

Introducing Tim — your cloud on autopilot:

Tim, the Taloflow Instance Manager, is an autopilot for cloud resource management that saves you up to 40% on AWS. Tim automates the discovery and cost tagging of all your resources on AWS. Tim works in real-time and through its AI model can calculate your AWS spend by the minute. Tim monitors all your resources and workloads and provides tailored, actionable recommendations on cost savings through Slack.

How does it work?:

  1. Automated discovery and cost tags for all AWS resources

Tim automatically discovers all your AWS resources and auto-tags them with appropriate cost tags. You can easily filter and apply tags to all your resources to ensure good tagging hygiene and analyze cost-per-hour spend on particular resources and tag groups. Through the UI you can apply instant actions such as stopping, napping, or terminating a resource.

2. Real-time AWS cost monitoring through AI cost model

Tim has a proprietary embedded AI cost model of all AWS resources and can accurately provide you with real-time insight and data on what your AWS spend is by the minute. This allows you to quickly see any anomalies on your cloud spend and also get real-time feedback on spend changes.

Tim also monitors a new metric called engagement which is a rules-driven analysis of your resources to see whether or not a certain is being used. This not only factors in activity but also considers other inputs such as who the owner of the resource is, what type of workload is running on that resource, and frequency of use.

3. Tailored, actionable recommendations on cost savings

Recommendation from Slack

Tim monitors your resources and workloads in real-time and provides timely recommendations to optimize your cloud spend and delivers these recommendations through Slack. You can directly stop, terminate, or schedule a resource to be restarted at a certain time (nap feature) without having to open up the UI.

We’re releasing Tim as a public beta. Please click here to see a demo with live AWS data and apply to become a beta user today. The beta program is completely free to take part in and can save you a lot of money on your AWS bill.

Going to AWS Reinvent next week? I’d love to chat and give you an in-person demo of Tim. Shoot me an email at jason@taloflow.ai.

--

--

Jason Kim

Head of Product @TaloHQ. Former Director of Product @poweredbygrow & growth at @wiivvit.