Anomalo
Your data is always talking. Don't miss what it's saying.
382 followers
Your data is always talking. Don't miss what it's saying.
382 followers
Your data changes constantly. Anomalo Analyst tells you what matters before you know to ask. It proactively monitors your Snowflake, Databricks, or BigQuery data, surfaces important trends, anomalies, and shifts, and lets you investigate with follow-up questions in plain language. Every insight is verified against your data, with Anomalo helping distinguish real business changes from broken data.
Free
Launch Team / Built With
Render Sandboxes: a computer for every agentIsolated compute, persistent state and a real filesystem
Promoted
Hey Product Hunt, I'm Elliot, co-founder and CEO of Anomalo.
I've spent years working with data — growth and product roles at @instacart and @LinkedIn, then building enterprise data quality tools at @Anomalo . I’ve seen that some of the biggest lessons often came from changes in the data. At Instacart, launching one new retailer surfaced an engagement spike that reshaped our growth strategy for years. At LinkedIn, small experiments in network-building tools moved user behavior enough to reset our roadmap.
But spotting those data changes is hard. You need dashboards for everything that matters and someone watching obsessively. Even with teams of analysts, we missed a lot.
That's why we built Analyst. Using technology Anomalo pioneered for data quality monitoring, it learns what's normal for your data, then uses AI to spot and explain meaningful changes — creating an automatic news feed of what moved and why.
Most data tools only answer the questions you thought to ask. The valuable data changes are often the ones nobody asked about: a paid social campaign that spiked purchases but only for certain SKUs, a supplier whose delivery times crept up 52% over a quarter, a metric that degraded only because of a mix shift. Anomalo Analyst spots these changes for you.
Just connect @DataBricks , BigQuery, or @Snowflake , read-only, and tell Analyst what matters in each table so the insights are tailored. You’ll start seeing Analyst insights in the first few days. Each insight includes a business summary, a technical summary, and the underlying logic and investigative steps.
Free to start. Give it a try and then tell us what it found.
Congratulations on the launch, it looks great. Does it require a certain amount of data or can it work for startups or sites with low volume?
@you_x_you_i thanks for the kind words! Analyst will work with small datasets too. The types of anomalies detected and level of confidence in the math change with higher volumes of data but as long as you have some fresh data coming in on a regular basis we should find insights. If you try it and find it's not catching what you expect, let us know and we can help.
@eshmu @john_joo2 @joka congrats on the launch ! How do you differentiate this from the already existing tools ?
@john_joo2 @joka @easeops Anomalo Analyst is proactive. You just set it up, and it delivers insights and interesting observations to you -- no need to ask it questions or prompt it in any way. That's a big difference from other AI analytical tools.
@john_joo2 @joka @eshmu That's crazy!
@john_joo2 @joka @easeops I like to say it's magical! :)
@odeth_negapatan1 Thanks for the support!
step 2 has AI agent decide what matters after statistical modeling ranks changes by magnitude. in practice painful work of framing question is where operators build mental model of the business. if tool decides what is worth noticing before human forms hypothesis, understanding of system atrophies or relevance just becomes whatever had highest statistical variance )
@konstantin_tikhaev Good point, but we've found our users to have the exact opposite response. They aren't just leaving the AI where it left off. Instead, they are naturally curious. For each generated insight, they often have followup questions and click on "Dive in with Analyst" where they can ask. In essence, AI doesn't end up cutting out the thinking. It ends up cutting out the manual labor of writing SQL or having to watch the dashboard all day.
@john_joo2 Reacting to an anomaly is still reactive debugging. You form the mental model by deciding which questions matter before the system flags them. Not by exploring the ones it chose)
@konstantin_tikhaev Very valid concern. What we've found is that Anomalo Analyst supplements that building of mental models, not outsources it. Analyst speeds up the exploratory process where you're exploring and framing those mental models. In the process it learns how you think of the business and it flags unexpected changes that you hadn't thought to look for - so you get alerted when something happens that you never predicted.
@joka catching unexpected anomalies gives "coverage", sure. friction of manually wrestling with data is what actually builds intuition, speeding past that struggle skips how mental model gets built
the "distinguish real business changes from broken data" line is the part that'd make or break this for us. we run a call-quality pipeline where a metric drop is sometimes a real regression and sometimes just a carrier reporting gap for a few hours, and naive anomaly detection flags both identically. curious how Analyst tells those apart in practice - is it mostly pattern-matching against known data-quality signatures (nulls, schema drift, volume drops) or does it also learn what a plausible business explanation looks like for a given table over time?
This makes data quality monitoring feel much more proactive and useful Love the focus on catching issues before they become bigger problems
@shivam_kushwaha16 Thanks -- that's exactly how we see it!
@shivam_kushwaha16 That's what we were aiming for! Just like going to the doctor, being proactive can save a ton of pain down the road. But, we're not just looking at data quality issues. We wanted to make sure we're highlighting opportunities in your data as well. After all, missing something in your data that tells you a way to 10x your business is likely much more costly than missing a typical data quality issue.
@john_joo2 got it