Asset Intelligence by Human Managed
Know what data you have. Know what just changed.
Most teams cannot answer either question from one place. The data sits in a warehouse, a lake, a pipeline tool, a model registry, and a glossary nobody has opened since last quarter. Asset Inventory is the catalog: sources, datasets, pipelines, models, and the terms that describe them. Asset State is the health record: quality, schema, freshness, volume, lineage. When a source goes quiet, you can see what else broke without opening five tools.
Most downstream failures start with something nobody was watching.
They rarely start with something exotic. They start with something ordinary that slipped out of view: a source that stopped landing, a field that went null, a pipeline that ran on yesterday’s schema. The data was there the whole time. It just changed while everyone was looking somewhere else. (Usually at another dashboard.)
The data usually exists. It is just scattered: a warehouse, a lake, a few SaaS exports, a model registry, and a glossary last touched in another quarter. Each holds a piece. None of them agree. Asset Inventory pulls all of it into one record per asset, and Asset State keeps watching after the inventory is built. You stop reconstructing the picture every Monday and start working from it.
TWO SIDES OF THE SAME PROBLEM
What you have, and what changed since.
Asset Inventory answers the first question. Asset State answers the second. Together they cover the two things digital operations teams get asked the most.
What data do we actually have?
Every data asset in one place, fresh enough to trust
Sources, datasets, pipelines, models and business terms, in one catalog. See the total count, how it breaks down by type, and when each record was last seen, so you know the view is current rather than a snapshot from last Thursday.
- One record per asset, de-duplicated across every tool
- A freshness indicator on every entry
- Flexible search and filter that narrows to a domain or owner in seconds
- Criticality and quality tracked as separate things
- Owner, lineage, classification and confidence score on every asset
What changed since someone last looked?
A health record for every data asset, updated continuously
Asset State tracks the condition of each data asset over time. It captures schema changes, quality movement, volume shifts, freshness breaches, lineage breaks and classification updates, then lays them out on a timeline you can read.
- Timeline at operational resolution, not a monthly report
- Drift detection against your last known-good baseline
- Contributing factors spelled out, with thresholds
- Continuous evidence for auditors, not a point-in-time check
- Change context: a null-rate spike on a revenue feed is not a new column on a sandbox extract
06:15 to 06:30 · Critical · volume 38% of 7-day baseline (threshold 10%) · settlement_ts null rate 41% (threshold 2%) · pipeline ran; payload incomplete
Every data asset type. One record each.
If it feeds a decision, it belongs here. You keep running the tools you already have, and controlshift.io makes them agree on one record per asset instead of five conflicting ones.
Data sources
Warehouses, lakes and SaaS systems, with connection health and last-seen tracked over time.
Datasets
Tables, files and streams, classified, owned, and scored for quality.
Pipelines
DataOps workflows, with freshness, failure and schema tracked as they run.
Models
Training sets, versions and drift, tied back to the data they depend on.
Business terms
Glossary entries mapped to real assets, so the language and the data agree.
Intelligence products
Catalogued outputs the Intel Engine actually serves, not a slide from last quarter.
Monday morning, a source goes quiet. What else just broke?
A real path through the platform, from question to answer in seconds instead of a morning of cross-referencing five catalogs.
Filter to what matters
A freshness alert lands on the customer-transaction source. Filter the catalog to that domain and it returns every dependent dataset, pipeline and model, ordered by business criticality. The two that feed reporting sit at the top.
Open the asset
Click in for the essentials: it is a production table in the finance warehouse, owned by Data Ops, classified confidential, last seen four hours past SLA. Overall state is red.
Ask why
Hit Explain. The AI walks through the classification: the red state comes from freshness and upstream lineage, even though quality scores are still green. The source landed three of four expected files. That missing land is the real story.
See what changed
Switch to Asset State and the timeline shows exactly when it turned. One window logged a 62% volume drop against a 10% threshold, and settlement_ts went null on a business-critical feed.
Act in place
Assign it to the right person, flag it for investigation, or acknowledge it as a planned migration. The decision gets recorded, so when someone asks who signed off, you have the answer.
When a state does not add up, ask it.
Every score, state and anomaly has an Explain button. The AI reasons over the asset’s actual data (its quality, freshness, lineage, volume and classification) and tells you why it landed where it did, in plain language, and what you can do next.
It reads your business context too. A well-scored, correctly classified feed that stopped landing on time gets treated as worth investigating now, instead of being buried under assets that only look messy on paper.
The picture stays current on its own.
Inventory and state do not sit in a corner. They feed the rest of controlshift.io automatically, and you just see the result.
Intelligence recalculates automatically
When an asset’s state changes, the Intel Engine reruns on current data. A feed that read reliable yesterday can read degraded today because volume dropped or a field went null. You do not have to reassess it by hand.
Models see the data they actually trained on
When Model Ops watches for drift, it builds on what is in the catalog right now, not on last quarter’s schema or an assumed training set.
Decisions on the record
Assign, flag or acknowledge a change in place. Every call is logged, which is exactly the evidence operations, risk and auditors are looking for.
Frequently Asked Questions
You can keep governing from last quarter’s catalog. Or you can work from what is actually there.
Asset Inventory and Asset State are part of controlshift.io by Human Managed.