Question 1

What data sources can the AI KPI Performance tool connect to?

Accepted Answer

It connects to more than 100 data sources, including Snowflake, BigQuery, PostgreSQL, Salesforce, Google Analytics, and custom REST APIs, which covers most of what a mid-size to enterprise company already has spread across its data warehouse, CRM, and marketing stack. The Connect Data Sources step is the first thing we do during setup, before defining any KPIs, because the causal inference modeling and anomaly detection only work well once the platform can see data across systems rather than just one isolated source. Companies using Tableau or Power BI already don't need to migrate off them either, since the platform can push insights back into those dashboards instead of replacing them.

Question 2

How quickly does it detect anomalies?

Accepted Answer

Under a minute, typically. The system processes incoming data streams continuously, not on a batch schedule, so a statistically unusual pattern, like a sudden spend spike or a metric drifting outside its normal range, gets flagged and routed to the right stakeholder almost as soon as it happens rather than showing up in a weekly report after the fact. This matters most for operational efficiency tracking and marketing ROI attribution, where catching a problem within the hour can mean the difference between pausing an underperforming campaign early and burning a week's budget on it. The anomaly detection runs alongside the real-time dashboards, so the same view executives already check surfaces the alert too.

Question 3

Do I need a data science team to set up the predictive forecasting?

Accepted Answer

No. The predictive forecasting feature uses automated machine learning to pick the best-fitting model for your specific time-series data, so there's no coding or manual model selection required from your team. This is deliberate: most companies asking for sales pipeline forecasting or executive dashboards don't have a dedicated data science team on staff, and we didn't want that to be a prerequisite for getting value out of the causal inference modeling underneath. You define the KPI you care about during the Define KPIs step, and the AI handles picking and tuning the forecasting approach itself during the AI Analyzes stage. A data science team can still plug in custom models via the API if they want more control.

Question 4

Can I set up custom alerts for specific KPI thresholds?

Accepted Answer

Yes, you can configure conditional alert rules that combine multiple metrics, like being notified if revenue drops 10% week-over-week while ad spend is simultaneously increasing, rather than only getting a flat single-metric threshold alert. This is where the causal relationship discovery feature adds real value over a basic dashboard: instead of two separate alerts that don't obviously connect, you get one alert describing a specific combination of conditions that's actually meaningful to your business. Marketing ROI attribution and operational efficiency tracking both lean on this heavily, since the interesting problems in those use cases are usually two or three metrics moving against each other, not a single number crossing a line.

Question 5

How are the executive reports generated?

Accepted Answer

The AI uses natural language generation to turn the dashboard data into a readable summary, calling out the key wins, the risks worth attention, and specific recommended actions rather than just handing over a wall of charts. This is the Custom Reports feature, and it's built specifically for the C-suite executive dashboards use case, where an executive usually wants the three-sentence version before diving into the underlying numbers, if they dive in at all. Reports pull directly from whatever KPIs you defined during setup, so the language reflects your actual goal-tracking targets, not generic commentary, and they can be scheduled to generate automatically on whatever cadence your leadership team reviews performance.

Question 6

Is my proprietary business data secure?

Accepted Answer

Yes, data is encrypted end-to-end, both in transit between your connected data sources and at rest inside the platform. For enterprise clients with stricter security requirements, we also offer single-tenant and on-premise deployment options, so your commercial metrics and causal inference models never sit on shared infrastructure with other clients if that's a hard requirement for you. This matters more than it might for a typical analytics tool, since the platform is doing deep regression testing on raw commercial data to isolate what's actually driving profitability, which is exactly the kind of sensitive competitive information companies don't want exposed. Data source credentials are stored separately from the analytical data itself as an added layer.

Question 7

Can it integrate with our existing BI tools like Tableau?

Accepted Answer

It runs as a headless intelligence layer that pushes predictive insights and anomaly flags directly into your existing Tableau or Power BI dashboards, so your team never has to abandon tools they already know for a whole new interface. For a lot of enterprise clients this ends up being the preferred setup: the causal inference modeling and predictive forecasting happen inside the platform, but the actual viewing experience for most employees stays inside the BI tool they're already trained on. Data Integration works the same direction too, since Tableau and Power BI connections are two-way rather than the platform being a one-directional data source.

AI agent case studies

Explore all 11 AI agents

Enterprise Conversational AI: The 2026 Guide

Book a free AI audit

Enterprise Conversational AI: The 2026 Guide

AI KPI Performance

What is AI KPI Performance?

AI Decision Intelligence

Everything you need. Nothing you don't.

Real-Time Dashboards

Anomaly Detection

Causal Relationship Discovery

Predictive Forecasting

Goal Tracking

Custom Reports

Data Integration

Without AI

With AI KPI Performance

Connect Data Sources

Define KPIs

AI Analyzes

Act on Insights

More agents in the suite

AI Procurement Agent

AI HR Agent

AI Cold Email

The services behind this agent

AI Recommendation Systems

Artificial Intelligence & Automation

Design Systems Development

UI/UX Design

FAQ: AI KPI Performance

Let's build your AI system