Subscribe
Sign in
Orbiter AI — Detect business issues without staring at a dashboard
Orbiter analyzes data in real-time to detect abnormal drops in business and product metrics. When a problem is detected, we alert teams in Slack so they never miss an issue that is impacting revenue or users.
65
Replies
Best
@markstevenwai @zhangwins @liveink Origin stories eh, where do I begin 😂Mark and I worked at Tesla's on the growth team from pre-Model 3 days till the Cybertruck unveiling. The pain of not knowing what was broken at any given time was a huge pain. Tesla moved incredibly fast - there were dozens of features being shipped each week, new sales updates being launched and tweaked, and changes to plans across all parts of the org. That meant your features might be accidentally impacted when another team ships code, launches a new email campaign, tweaks the website, ... the list goes on. Elon is also just really good at finding out what's broken via Twitter. Add the two together, and you can imagine the kinds of fires that break out when you see that Elon email rolling through before anyone had caught a graph trend dipping. Winston also has similar stories from Doordash too - it's what happens when you have growing companies that are trying to push the bounds on responsiveness and attention to customers.
@markstevenwai @zhangwins @victor_zhang1 Thank you! This line is amazing "Elon is also just really good at finding out what's broken via Twitter."
Yessss 🙏 Been using orbiter for quite some time now — so happy to see you launch! Amazing team. I signed up for peace of mind, frantically checking my dashboards every 10 minutes was just unhealthy. What's cool is it also notifies you in the opposite direction too, when things start to grow :)
@jakemor Thanks Jake! We're actually so pumped to solve this problem since we experienced it first-hand. Hard to imagine a world 5 years from now where metrics and dashboards aren't fully autonomous, monitoring themselves, and letting the user know when there's an insight (good, or bad)
So far our team has had an amazing experience with Orbiter -- such an important product, it's pretty crazy that there's so much monitoring in place for eng related metrics, but, previously, none for business/product metrics.
What other alerting integrations are you thinking of aside from Slack?
thanks @garymlin ! We're definitely excited to work on a pain point close to our hearts. We currently support Slack in Beta right now, but we are listening to our early users and considering building integrations for SMS, VictorOps, and possibly more to help teams receive their urgent alerts ASAP.
?makers Seems like a powerful tool. Would a future version of Orbiter build a model to demonstrate relationships between metrics? Paired metrics, positively or negatively correlated metrics (perhaps as a tool to identify misconfigured metrics). Maybe a better way to ask the question: what are you excited about beyond monitoring, what's the vision?
Hey @jacobhartog, we find it really hard to imagine a future where teams are expected to stare at dashboards and manually repeat adhoc analyses to detect issues or iterate on products. With Orbiter, we want to automate the entire workflow of problem detection and root cause analysis using software to learn from your workflows and historical data. It starts with high signal alerts that find real noteworthy trend changes to investigate, paired with the proper context to help you get to the root cause. Automatically surfacing correlated metrics (or dimensions) is a great example of that, in addition to ideas we have such as the ability to quickly adapt your previous workflows for similar analyses. In the future as we learn more, we’d like to be the product that helps you gets from the question about a trend to the actionable insights in the shortest amount of time.
Great to see it launch!! Known the the maker team for a long time and was an early beta user. It has really given us the time back to focus on other parts of the business.
@harry_chen Thanks Harry - awesome to hear! One of our product values is speed -- speed of detection, speed of diagnosis, speed of resolution -- and so we want to automate time-consuming analytics work so that companies have the ability to focus on higher leverage tasks
@vinojeya Orbiter currently connects to your analytics databases to set up monitoring and alerting. With GA, we support BigQuery where you can pipe your analytics data into! Similarly with Shopify, we currently support customers who pipe Shopify sales data to databases such as Postgres, MySQL, etc. (usually for further analyses). But we would love to support more teams so we are investigating support for their API connection for easier setup! That's a great question though, so I'd be happy to chat more about specific use cases if you'd like!
Cool product! Can't wait to try it out. I'm curious about real time metrics, is there plan to support those? Like data coming from kafka or non relational db's like nosql? Some of our real time metrics are fed/ingested through these. Also, what if i wanted to make or compute my own metric coming from different sources. Thanks.
@matthew_liem awesome questions, appreciate it! Currently we support streaming data into analytics databases in Postgres, where some of customers actually get data posted every minute! We currently do not have connections to Kafka or NoSQL implementations but we're always interested to get more users onboard! Please send us an email and I'd love to chat more about your specific implementation.
@matthew_liem in terms of computing metrics from different sources, we support this! That's our main benefit of staying platform/database agnostic to enable really interesting joins across multiple databases. For example, if you had revenue in one db, and geographic order data in another, you could get automated monitoring and alerting for your revenue by region!
Great product! At first glance it seems super useful, definitely looking forward to try it out! It was great meeting at Startup Grind!
Hello 👋 Product Hunt! We are Victor, Mark, and Winston, founders of Orbiter.
Before Orbiter, we were product managers and data scientists at Tesla, DoorDash, and Facebook. It often felt impossible trying to keep up with the different dashboards and metrics while also actually doing work and building things. Even with tools like Amplitude, Tableau, and Google Data Studio, we would still catch real issues late by days or weeks. We saw that our engineering counterparts had plenty of tools for passive monitoring and alerting -- PagerDuty, Sentry, DataDog, etc. -- but the business and product side was lacking in solutions. We built Orbiter to solve this.
Orbiter builds machine learning models for your metrics that capture the normal/abnormal patterns in the data. We use a supervised learning approach for our alerting algorithm to identify real abnormalities. Orbiter accomplishes this by forecasting the expected “normal” metric value and also classifying whether an abnormality should be labeled as an alert.
Orbiter is easy for non-technical teams to set-up and use. It’s a Web app, requires no eng development, and connects to existing analytics databases the same way that existing dashboard tools like Looker or a SQL editor just plug in. Teams connect their Slack to Orbiter so they get immediate notifications when a metric changes abnormally.
Here’s an Orbiter use case example: for an e-commerce app, a number of endpoints were migrated in Q4 last year which unknowingly caused a feature in the Android shopping flow to disappear. Typically, users in that part of the shopping flow progress to the next page at a 70% rate but because of the missing feature, this rate dropped by 5% absolute. This was a serious issue but was hard to catch by looking at dashboards because: 1) this was just one number changing out of hundreds of metrics that change every hour, 2) this number naturally fluctuates daily and weekly, especially as the business grows, 3) it would have taken hours of historical data analysis to ascertain that a 5% drop was highly abnormal for that day. It wasn’t until this metric stayed depressed for many days that someone found it suspicious enough to investigate. All in, including the time to implement and deploy the fix, conversion was depressed for seven days costing more than $50K in reduced sales.
We’d love to hear your feedback! Please let us know below or feel free to send us a note at hello@getorbiter.com. Thank you!
Hmm looks super dope, but seriously, does it work? I've built manual alerting on my product eng team and we get a lot of noise :(
@mskpw Hey Steven - the accuracy of alerting is really important to us and is something we work closely with our customers on. for each metric, we build a custom ML model that is adjusted to meet your needs (e.g. time to detection, sensitivity, etc.). happy to chat about it more! feel free to DM me at winston[at]getorbiter.com :)
@zhangwins yea I feel like accuracy is a huge pain point for me. If you can do it, then I'm all in haha! Super interested to hear more, I just sent you an email!
Really good product - it's sometimes just too much effort constantly monitoring significant changes to your metric, this is a life-saver!
@grisha_gevorkyan Thanks Grisha! That was exactly our motivation for building Orbiter. Having to track pages and pages and pages of dashboards is so time-consuming in today's fast moving world.
Is Orbiter only for businesses of a certain size? We’re a fresh and new business, so wondering if this is something I should wait on to implement..
Hey @abigail_hung we currently have customers both large and small! For our startup customers with smaller data sets, we work with them to implement thresholds such as "alert me when metric falls below X value" or "alert me when value drops Y% vs last week". It's a hugely important use case to use as we're a startup too! Send me a message at mark[at]getorbiter.com and I'd love to chat about specifics!
@emily_wai Orbiter was a brainstorm to evoke the thought of a machine that's 2km up and has a birds-eye view on all your metrics to be able to see the full picture :D
Man this thing will save me / the team so much time.
As a data scientist, I spend so much time manually monitoring what's good, what's bad and what's ok and sometimes it's not even obvious to me if something is noise or actually a real trend. We tried using static 'threshold' filters in persicope or other tools but they have too many false positives. Seems like you guys have thought about how to trade off between false positives and false negatives, which is v clever.
Gonna give this a spin today!
@sumanyu_sharma thanks! Excited to hear this your thoughts and feedback on this as it sounds like you’ve thought about this place problem space! Our signal to noise ratio is one of our highest priorities, so I’m happy to chat more on that if you’d like! Please send us an email and we’ll get you set up :)
I can think of several instances when Orbiter would’ve proactively helped my team in the past and can see it becoming a necessity for any team looking to stay ahead of the game moving forward. Looking forward to following the progress and success of this tool!
@howie_young thank you! What dashboard or analytics tools are you using today? Happy to chat more about our pricing plan if you send me a note (it's winston[at]getorbiter.com!). Our pricing plan is a monthly subscription per seat (similar model as Looker) and we also have an enterprise plan for larger customers.
Such a great concept for a product - we're constantly trying to free up time by automating analytics against our KPIs. It's always a win for us when our business leads can spend more of their time layering in their contextual knowledge on top of the data vs. spending the majority of time gathering the data to begin with.
@chester_lau Thanks Chester! We're looking to help every company achieve operational excellence by better focusing time on building great products
We've been using Orbiter and the best part is getting the daily reminder that everything is running fine and we're on a daily streak w/o issues 👌🏻 Setting up was super easy, too
@jennelle_nystrom keep that streak going strong 💪 super happy to be working and helping amazing teams like yours!
Amazing! Metrics need to be monitored automatically, could totally see this product taking over Excel sheets, static BI dashboards, and even long-form presentations. This would handle so much of the hassle around tracking metrics as a PM, eng manager, or any data-driven job.
@aditya_vasanth_challapally Hey Aditya - definitely, there are a ton of different applications for monitoring and we'll have to be deliberate about the use cases to prioritize! I also imagine that as our core competency in abnormality modeling / predictions improve, there will be plenty of applications beyond the business/product cases as well
This solves a real pain for us running an enterprise SaaS - empowers us to "go on offense".
@keegan_huang That's right! part our long term vision is to enable teams to be more efficient so they can go more on offense.
Farmstead