256 karma · joined March 12, 2018
kaveh@usage.ai
Whoever wins these contracts, I hope that taxpayers get a good deal - not that we have much choice in the matter. For example, my company, Usage.AI, cuts AWS bills by up to 57% using Reserved Instances, but without any lock-in.
At the scale of $10 billion, every percentage point cut is worth $100 million. With the average business wasting 35% on their AWS bill, that's $3.5 billion of potential savings that could be going back into the taxpayer's pockets (https://www.flexera.com/blog/cloud/aws-costs-how-much-are-yo...).
But I digress.
AWS is clearly the leader in cloud computing, and they have been for some time. They have made significant investments in their infrastructure and their team, and it shows.
The same applies to cloud costs more broadly, across providers. With Reserved Instances, for example, AWS dramatically cuts the cost of running an instance if you commit to using it for a one- or three-year term. But signing up for a Reserved Instance requires a lock-in commitment, so many people don't bother because they're worried about being stuck with an unnecessary cost if their usage changes. My company, Usage.AI, created a solution: We cut EC2 costs by up to 57% using Reserved Instances, but without any lock-in. If an RI is no longer needed, it's just sold on the open market. No hassle, no risk.
If the results are anything like what other reports on the topic say, then businesses are finding it hard to predict their cloud costs, they're spending too much on cloud services, and they're struggling to control costs [1]. All of which are problems that need to be addressed - my company, Usage.AI, enables businesses to cut AWS costs by up to 57% using Reserved Instances, but without lock-in [2]
[1] https://www.forbes.com/sites/forbestechcouncil/2022/04/19/cl... [2] https://www.usage.ai
That's why a lot of businesses are going the "my own computer" (on-premises) route, but there are actually a ton of advantages to the "someone else's computer" (cloud) route.
Beyond the obvious (scale, speed, uptime, redundancy, features), you get increased collaboration, easier upgrading, and a lot less code to manage. My company, Usage.AI, uses Amazon's Reserved Instances marketplace to cut customers' AWS bills by up to 57%, without any lock-in.
However, for larger projects, the cloud offers a ton of advantages. There's the obvious ease of collaboration, but you also get access to a lot of powerful tools and resources that can make development a lot smoother. Plus, you can easily scale up or down as needed, which is great for resource management.
You also eliminate large upfront CapEx costs, but the trade-off is often surprisingly high OpEx costs. In the end, it really depends on what you need and what your budget looks like.
If you go with the cloud, my company, Usage.AI, takes advantage of the Amazon Reserved Instances marketplace and usage analysis to buy and sell RIs for you, cutting up to 57% off of your AWS bill. Check us out at https://www.usage.ai!
That's why my company, Usage.AI, has built a tool to automatically buy and sell Reserved Instances on AWS to cut costs. We're not on GCP or Azure yet, but businesses using us cut up to 57% on EC2 costs.
Looking at YCombinator's alumni, there are a few interesting ones - for example, CloudThread.io (no affiliation) tracks cost efficiency of teams and applications, providing "technical cloud cost unit metrics as a service."
My company, Usage.AI, automatically buys and sells Reserved Instances to cut EC2 costs by up to 57%.
Mindshare is definitely key, but there's also the fact that AWS has been around longer and has a more robust services offering. And as the first mover, AWS has been able to set the market standard for pricing.
That's a double-edged sword though, and many businesses are dramatically overpaying for AWS services [2].
On a related note, I built Usage.AI [3] to automatically buy and sell Reserved Instances on AWS to cut costs.
[1] https://www.engadget.com/amazon-web-services-re-awarded-nsa-... [2] https://www.forbes.com/sites/forbestechcouncil/2022/04/19/cl... [3] https://usage.ai
My company, Usage.AI, has built a tool to automatically buy and sell Reserved Instances to cut EC2 costs by up to 57%.
I'm sure the NSA got some sweet discounts on a contract this big, but I always wonder how much these agencies are spending on the cloud overall. Maybe they're overpaying in the billions of dollars?
That triopoly - or oligopoly, if you prefer - is really bad for startups.
It's not just that the big players can charge whatever they want. They also keep coming up with new ways to raise prices.
It's gotten to the point where startups are at risk of bankruptcy because of unexpected cloud costs [1]. Given their dominance, it's no surprise that Google recently hiked prices [2].
There are many ways to cut cloud costs, but founders I've spoken to don't have the time to optimize their spend. On a related note, my company, Usage.AI, has built a tool to automatically buy and sell Reserved Instances to cut EC2 costs.
If you're a startup on a tight budget, you can't afford to overspend on the cloud. And given the big players' continued price hikes, it's only going to get harder to stay within your budget.
[1] https://www.forbes.com/sites/forbestechcouncil/2022/04/19/cl... [2] https://cloud.google.com/blog/products/infrastructure/update...
kaveh@usage.ai
So you should actually buy the RI for the instance you plan on shutting off, and then we will buy it back when you shut the instance off.
Feel free to ping me directly if you'd like: kaveh@usage.ai
Just my 2 cents: Would be amazing if this could automatically change to the most cost-effective machine type for you, so the user wouldn't have to go in and do it manually. Perhaps that's something you're already working on?
On a related note, I built Usage.AI [2] to automatically buy and sell Reserved Instances on AWS to cut costs.
[1] https://techcrunch.com/2022/03/14/inflation-is-real-google-c... [2] https://usage.ai
Kaveh here, founder of www.usage.ai
I'm organizing a Webinar with the goal of companies being able to further optimize their own AWS spend. In this webinar, we will be going over 21 different ways you can reduce your AWS spend.
We use the sts:AssumeRole policy to create temporary short-lived credentials for us to get access to the AWS APIs on your behalf. The assume role permission is constrained only to the policy we've defined on our landing page and in our app, which are read-only + the ability for us to manage your reservations on your behalf.
Our RIs are actually more flexible than SP. There is no commitment and if you want to change region or instance type, Usage will buy the old type and sell you the new type.
We chose RIs because AWS allows us to buy and sell RIs. There is no marketplace for SPs at the moment.
Looking forward to answer your questions!