The complete guide · AI Analytics and Forecasting

AI Analytics and Forecasting: the complete guide

How decision intelligence actually works, from a plain-language question through the semantic metrics layer to a grounded answer, with anomaly detection and forecasting with honest uncertainty. Written so you understand the engine before you ever share a byte of data with us.

A working reference, not a sales brochure. When you want it run on your data, start with a free analysis.

For years, getting an answer from your data required a data analyst, a ticket queue, and a few days of back and forth. The tooling has changed enough that none of that is necessary. What has not changed is the discipline required to make the answers trustworthy: you cannot just point a language model at a raw database and call it analytics. This guide is the whole method, written out.

What decision intelligence is

Decision intelligence is the practice of making business data answerable in plain language, with the answers grounded in agreed definitions so they mean the same thing to everyone who asks. It is a layer that sits on top of your data warehouse and turns it from a place where data is stored into a place where decisions are made.

It helps to separate three things that often get collapsed into one pitch:

  • Self-serve analytics is the ability for a business user to ask a data question and get an answer without filing a ticket. This is the most visible capability, but it depends entirely on what is underneath it.
  • Anomaly detection is the system watching your metrics continuously and alerting you when something breaks its expected pattern, with enough context to start an investigation rather than just a number that is red.
  • Forecasting is a model trained on your historical data that projects forward with an explicit uncertainty range, so you can plan with the range in view rather than anchoring on a single number that will be wrong.

Each of these can stand alone, but they compound. A team that can ask any question, gets alerted when something breaks, and has a forecast for what is coming operates fundamentally differently from one that waits for the weekly review.

The free analysis

Every engagement starts with a free analysis, because it is the only honest way to scope the work. You share a sample of your data and a list of the questions your team is currently asking a person to answer. We model the metrics, run the analysis, and come back with real answers grounded in your numbers, anomalies surfaced on your history, and a forecast with its accuracy measured.

You keep the result whether or not you continue. A measured analysis on your real data is a useful artifact on its own, and it ends the guessing that kills most data projects before they start. If the numbers hold up and the answers are genuinely useful, we scope the full build from there.

Natural language to SQL

The interface that most people see first is the question box. You type "why did revenue dip in March?" in plain English and an answer comes back. The mechanism behind that is a language model that translates your question into SQL, runs it against your data, and returns the result with a chart.

This is where most demos stop and where most projects fail. A language model translating a free-form question into SQL against an undocumented, unlabeled database will produce SQL that runs but returns the wrong number with high confidence. It will join the wrong tables, apply the wrong date filter, and miss the business logic that makes "revenue" mean gross or net depending on the context. The output looks plausible. It is not trustworthy.

The fix is not a smarter model. It is a semantic metrics layer, which is the next section.

From a plain-language question to a grounded answer
QUESTION "why did revenue dip?" SEMANTIC METRICS LAYER defined metrics, logic, access rules, grain SQL generated + validated WAREHOUSE your data, your tables GROUNDED ANSWER result + chart + anomaly if found
The semantic metrics layer is the critical middle step. Without it, the model writes SQL against raw tables and the answer depends on guessing the schema. With it, the model writes SQL against defined, tested metrics and every number is traceable to a business definition.

The semantic metrics layer

The semantic metrics layer is the set of definitions that sit between your raw tables and the question interface. It is where "revenue" gets defined as orders.amount minus refunds.amount, filtered to completed orders, with the rule that a refund applied to a prior-period order adjusts the current period rather than the original one. That definition is encoded once, tested, and then every query uses it.

Building this layer is the majority of the engagement work, and it is also where most of the value sits. Once the metrics are defined, a business user asking "what was revenue in March?" and a data analyst writing the same query will get the same number, because they are both running against the same definition. That consistency is what makes the answers trustworthy enough to put in an exec brief or a board deck.

The layer also carries grain (is this metric daily, monthly, by region, by product?), access rules (who can see which metrics?), and dimension mappings (when someone says "UK", which column values does that map to?). These are the details that a raw language-model query will get wrong, and the layer is what makes getting them right reliable rather than lucky.

Grounding answers in defined metrics

Grounding means that every number in an answer can be traced back to the SQL that produced it, and that SQL can be traced back to the metric definition that informed it. A user who questions a number can see exactly how it was computed, which column it came from, and which filter was applied. That is the difference between a result you trust and a result you verify every time.

This matters most at the edges: when a forecast looks surprising, when an anomaly alert fires, when an exec challenges a number in a brief. In each case, the answer to "how did you get this?" is a clean, auditable chain from definition to query to result. That chain is what earns organizational trust in the system over time.

Anomaly detection

Anomaly detection is the system watching your metrics continuously and alerting when a value breaks outside its expected range. The alert includes the affected metric, the dimension that is driving the break, the magnitude of the deviation, and a plain-language description of the likely cause based on historical patterns.

The engineering behind this is straightforward in concept: for each metric, we fit a model of normal behavior on your historical data, set alert thresholds based on your tolerance for false positives, and then compare each new observation against that model. In practice, the interesting work is in the dimension analysis: not just "revenue is down" but "revenue is down and it is entirely explained by a spike in refund rate in the UK, which has happened three times before when a specific product category had a quality issue."

Anomaly detection: a metric line with a flagged spike
Anomaly flagged Revenue down 24% UK refund rate: 18% vs 4% JanFebMar AprMayJun JulAugSep OctNovDec Revenue Expected range Actual metric Anomaly
The alert fires before the Monday review. It includes the dimension driving the break (UK, refund rate) so the investigation starts with a direction rather than a blank screen. The expected range is fitted on your own historical data, not a generic industry curve.

False positive rate matters. An alert system that fires every other day trains people to ignore it. We calibrate the thresholds during the Build phase using your historical data and your stated tolerance, and we revisit them during Operate as the business changes.

Forecasting with honest uncertainty

A forecast is a claim about the future, and honest uncertainty is the most useful thing a forecast can contain. A single-point prediction ("revenue next quarter will be EUR 2.4M") is less useful than a range ("the model puts 80% of the probability mass between EUR 2.1M and EUR 2.7M, based on this backtest error"). The range is the thing you plan around.

Our forecasts are built on three disciplines:

  • Backtesting first. Before any forecast goes into production, it is run on historical data as if it had been made at each past point. The measured error on your own data tells you what to expect. We share that number before you rely on the forecast for anything.
  • Explicit confidence bands. Every forecast output includes a lower and upper bound at a stated confidence level. Where uncertainty is genuinely high, the bands are wide, and the system says so rather than narrowing them artificially.
  • Model selection per metric. Revenue seasonality, inventory demand, and customer churn have different statistical signatures. We fit the model that matches the signal rather than applying one method to everything.
Forecast with an 80% confidence band
Actual (12 months) Forecast (6 months) JanFebMar AprMayJun JulAugSep OctNovDec Revenue Actual Forecast 80% confidence band
The band widens as the forecast extends further out, which is the honest representation of how uncertainty compounds over time. A model that shows a narrow band six months out is overfitting or overclaiming. We flag both cases in the review before any forecast goes into production.

What changes the accuracy of a forecast is the quality of the input data and the stability of the underlying process, not the sophistication of the model. A business that changes its pricing strategy every quarter will have a harder time forecasting revenue than one with a stable pricing structure. We make that plain during the analysis rather than promising accuracy we cannot deliver.

On backtesting. The backtest figure we show you in the free analysis is the mean absolute percentage error (MAPE) of the model on held-out periods of your historical data. It is the number you should use to judge whether the forecast is useful for your planning horizon, not a figure we chose to make the pitch look good.

Automated narratives and exec briefs

Once the metrics layer is in place and the queries are running, generating a weekly or monthly executive brief is a composition step. The system pulls the key metrics, compares them to the prior period and the forecast, surfaces the anomalies that fired, and writes a short narrative that describes what happened, what to watch, and what the forecast shows. It is sent on a schedule or triggered on demand.

The brief is not creative writing. It is a structured summary of the data, written in plain language, with the numbers grounded in the same definitions that answer the interactive queries. A manager can read it in two minutes and walk into the leadership review knowing the numbers without having built a slide deck.

The value here is not that the narrative is beautifully written. It is that it is ready at 07:00 on Monday rather than at 11:00 after someone has spent the morning pulling it together. The time that frees up is real and it compounds across every week the system runs.

Where it connects

The system sits on top of your existing data warehouse rather than replacing it. We connect to BigQuery, Snowflake, and Postgres as primary targets, with Redshift and Databricks available for teams already running on those platforms. The semantic metrics layer is the portable artifact: it can be defined once and queried from the natural-language interface, from a BI tool like Looker or Metabase via their API, or directly via SQL for teams that want to blend both approaches.

LayerWhat it doesWhat you keep
Warehouse (yours)Stores the raw and transformed dataFull ownership, no migration
Semantic metrics layerDefines metrics, grain, access, dimensionsPortable YAML or SQL definitions
Query engineTranslates questions to SQL, runs them, returns resultsAPI access for embedding elsewhere
Alert and forecast serviceWatches metrics, fires anomaly alerts, runs forecastsConfigured alert rules and thresholds
Brief generatorComposes and distributes the exec brief on scheduleBrief templates you own and edit

Nothing in this stack requires you to move your data. The query layer connects to your warehouse with read-only credentials. The data stays where it is.

Governance and access control

Access control is not an afterthought in a well-built metrics layer, it is part of the schema. Each metric is tagged with the roles that are permitted to see it. The query layer checks those tags at runtime: a user asking about executive compensation or customer PII who is not in the right role gets a response that tells them the metric exists but is not available to them, rather than an error or a data leak.

The governance model we recommend is simple: start with the broadest access that is safe, measure which queries are being run and by whom, and tighten where the data shows usage that was not intended. Overly restrictive access from day one kills adoption before the system has a chance to prove its value.

For companies under GDPR or other data-residency obligations, the read-only warehouse connection and the absence of a data copy in our infrastructure are the key facts. We document the data flow in the engagement and can provide the technical architecture for a compliance review.

Measurement and adoption

A system that nobody uses is not a success, however good the underlying technology. We track two categories of metric through the Operate phase:

MetricWhat it tells you
Time to insightHow long from a question to an answer, against the pre-system baseline
Query volume by teamWhether the system is being used and by whom, as a leading indicator of adoption
Forecast MAPE by metricHow the forecast accuracy evolves as the model is retrained on more data
Alert precisionThe share of anomaly alerts that led to a real investigation (true positives)
Brief open rateWhether the exec briefs are being read, as a proxy for whether they are useful

The adoption metric matters as much as the accuracy metric. A technically excellent system that is bypassed in favour of a spreadsheet has not delivered value. We revisit both categories in the monthly Operate report and adjust the configuration when either is trending in the wrong direction.

What does not work

A short list of approaches that look reasonable and fail in practice:

  • Pointing a chatbot at a raw database. This is the most common failure mode. Without a semantic metrics layer, the language model writes SQL against raw tables and the answers are plausible but wrong. The only way to catch the errors is to check every answer against a source of truth, which defeats the purpose.
  • Defining metrics during a sprint and never revisiting them. Metrics that were correct at launch become wrong as the business changes. A metrics layer without a governance process for updating definitions is a source of stale answers dressed up as fast ones.
  • Forecasting with too little history. A model trained on six months of data that includes a launch, a pricing change, and a seasonal peak has not seen enough normal behavior to generalize. We will tell you this in the free analysis rather than building a forecast that performs well in the demo and poorly in production.
  • Over-alerting. An anomaly system that fires on every small deviation trains people to ignore it. Threshold calibration is ongoing work, not a one-time configuration. We manage this through the Operate phase.
  • Treating the exec brief as a substitute for analysis. The brief is a summary of what happened. It is not a diagnosis of why it happened, and it is not a plan for what to do next. Teams that use it well use it as a starting point for a conversation, not as the conversation.

The engagement model

The shape of the work: a free analysis to prove the approach on your own data, a fixed-scope Build that connects the warehouse and ships the metrics layer, natural-language interface, anomaly alerts, and first exec brief, then an Operate phase that keeps the metrics and forecasts honest as your business and data change. You get one point of contact, software that runs on your infrastructure, and the measurement to prove it is working.

AI Analytics is our second AI service. The same methodology that makes Document AI trustworthy, start with a free pilot, measure before you commit, ship a real workflow, own the result, is what makes AI Analytics trustworthy. The services complement each other: Document AI turns documents into data, AI Analytics turns data into decisions. Start with a free analysis.


That is the whole engine. The next step is a free analysis on your own data: real answers to your real questions, the anomalies surfaced, the forecasts backtested, and no obligation.

Get a free analysis

Frequently asked questions

Will the AI invent numbers that are not in my data? +
No. Every answer is grounded in the semantic metrics layer, which is built from your own definitions. The system generates SQL against your real tables and returns only what the query finds. If a number is not in your data, it does not appear in the answer. This is the fundamental difference between a chatbot on a raw database and a decision intelligence system.
How is this different from a BI tool like Looker or Tableau? +
BI tools require someone to build a dashboard for every question in advance. This system answers questions that have not been anticipated, in plain language, without a dashboard to navigate. The semantic metrics layer is compatible with BI tools and can feed them, so you do not have to choose.
How long does the free analysis take? +
A few days from when you share the data sample. We come back with a 30-minute walkthrough of the results: real answers to your questions, the anomalies we found, and the forecast accuracy on your own history.
Do my data and queries stay private? +
Yes. The query layer connects to your warehouse with read-only credentials. We do not copy your data into our infrastructure. The data flow is documented in the engagement and available for a compliance review.
What does it cost to run on an ongoing basis? +
The Build is a fixed-scope engagement priced after the free analysis. Operate is a monthly retainer covering metric maintenance, model retraining, alert calibration, and the monthly accuracy report. We quote both numbers before any work begins, based on the scope the free analysis defines.

Ask your data the questions you have been waiting days to answer.

A free analysis on your real data, the metrics modeled and backtested, the anomalies surfaced. Then we build the layer that makes every answer take seconds.

Get a free analysis
Free analysis · You keep the result · 24h reply

Related reading

Related reading

Related reading

Get a free analysis