# Business Intelligence Agent

*/Agents/Business_Intelligence_Agent*

## Solution Overview

The Business Intelligence Agent connects directly to Snowflake or BigQuery to translate natural language queries into executable SQL and Python pandas operations. It operates as a dedicated data analyst for multi-channel retail operators, continuously monitoring inventory levels, ad spend, and SKU-level margins. The agent produces daily narrative briefings delivered via Slack and generates interactive, parameterized charts in response to ad-hoc executive questions.

Heads of Data and VPs of Operations hire this agent to clear the persistent backlog of routine data requests that typically stall strategic projects. Instead of waiting days for a human analyst to write queries and format tables, stakeholders ask the agent directly and receive validated answers in seconds. The agent flags anomalous data structures for human review, requiring a lead data engineer to approve any permanent alterations to the underlying dbt semantic definitions before they are published to the company dictionary.

Positioned as an active consumer of data infrastructure, the agent reads directly from headless ETL APIs and centralized data warehouses. Its outputs are consumed by human executives making immediate purchasing decisions and by downstream automated pricing services that adjust storefront markups based on the agent's real-time margin calculations.

## Icp Opportunities

- [Algorithmic Revenue Forecaster](/Opportunities/Algorithmic_Revenue_Forecaster) — Headless SaaS
- [AI Document Extraction](/Opportunities/AI_Document_Extraction) — Service-as-Software
- [Analytics Audit Agent](/Opportunities/Analytics_Audit_Agent) — Agent
- [Market Intelligence Service](/Opportunities/Market_Intelligence_Service) — Service-as-Software
- [Data Pipeline Orchestrator](/Opportunities/Data_Pipeline_Orchestrator) — Headless SaaS
- [Governance Enforcement Agent](/Opportunities/Governance_Enforcement_Agent) — Agent

## Agent Definition

**Goals**:
- [Query Resolution Time](/Metrics/Query_Resolution_Time)
- [Report First-Pass Accuracy](/Metrics/Report_First-Pass_Accuracy)
- [Data Request Backlog Volume](/Metrics/Data_Request_Backlog_Volume)
**Tools**:
- [Snowflake](/Products/Snowflake)
- [BigQuery](/Products/BigQuery)
- [dbt](/Products/dbt)
- [Slack](/Products/Slack)
**Skills**:
- [Data Analysis](/Skills/Data_Analysis)
- [Database Management](/Skills/Database_Management)
- [Statistical Programming](/Skills/Statistical_Programming)
- [Information Retrieval](/Skills/Information_Retrieval)
- [Critical Thinking](/Skills/Critical_Thinking)
**Contacts**:
- Slack
- API
**Identity**: did:web:agents.enterprise.example/business-intelligence-analyst
**Core Tasks**:
- [Translate Natural Language to SQL](/Tasks/Translate_Natural_Language_to_SQL)
- [Execute Python Pandas Operations](/Tasks/Execute_Python_Pandas_Operations)
- [Generate Interactive Charts](/Tasks/Generate_Interactive_Charts)
- [Calculate Real-Time Margins](/Tasks/Calculate_Real-Time_Margins)
**Escalation**: Any permanent alterations to the underlying dbt semantic definitions or flagged anomalous data structures escalate to a lead data engineer for approval.
**Memory Kind**: persistent
**Memory Note**: Retains prior query parameters, organizational dbt semantic definitions, and resolved data mapping rules to ensure consistent logic across recurring questions.
**Autonomy Mode**: guarded
**Replaces Role**: [Business Intelligence Analyst](/JobTypes/Business_Intelligence_Analyst)
**Solves Problem**: [Data Request Backlog](/Problems/Data_Request_Backlog)
**Responsibilities**:
- Monitor Inventory Levels
- Track SKU-Level Margins
- Answer Ad-Hoc Data Queries
- Distribute Narrative Briefings

## Agent Function Cascade

**Ai Role**: The AI acts as an end-to-end digital data analyst, independently receiving natural language queries, writing and iterating on SQL via tool loops, and generating analytical briefs, halting only to escalate permanent structural changes to a lead data engineer for approval.
**Cascade**:
- Kind: Generative · Note: Extracts intent and required metrics from natural language Slack inputs. · Step: Parse Ad-Hoc Request · Verb: parse · Realizes: Analyze Information Requests · Oversight: none
- Kind: Agentic · Note: Translates intent to SQL, iteratively querying Snowflake or BigQuery. · Step: Formulate And Execute Query · Verb: query · Realizes: Query Databases · Oversight: none
- Kind: Human · Note: Lead data engineer approves permanent changes to underlying dbt definitions. · Step: Review Schema Alterations · Verb: approve · Realizes: Maintain Database Dictionary · Oversight: approves
- Kind: Code · Note: Applies Pandas operations for margin logic and aggregations. · Step: Process Data And Calculate · Verb: calculate · Realizes: Analyze Business Operations Data · Oversight: none
- Kind: Generative · Note: Synthesizes data outputs into interactive charts and contextual briefings. · Step: Draft Narrative Briefing · Verb: draft · Realizes: Prepare Analytical Reports · Oversight: none
- Kind: Code · Note: Posts the final response back to the requesting user via Slack. · Step: Distribute Results · Verb: distribute · Realizes: Distribute Information · Oversight: none
**Optimizes**:
- [Query Resolution Time](/Metrics/Query_Resolution_Time)
- [Report First-Pass Accuracy](/Metrics/Report_First-Pass_Accuracy)
- [Data Request Backlog Volume](/Metrics/Data_Request_Backlog_Volume)

## Agent Representative Offer

**Warranty**: Guarantees 99.9% availability for the Slack interface and API, offering usage credits for queries that fail due to internal agent errors rather than underlying database timeouts.
**Price Band**: ~$0.15 to $0.75 per resolved data query, depending on query complexity and executed compute time
**Pricing Kind**: UsageMeter
**Deliverables**:
- Integrated Slack query interface
- Translated SQL execution logs
- Generated interactive charts
- Narrative inventory and margin briefings
**Delivery Mode**: The buyer provisions the agent directly into their Slack workspace and connects it to an existing data warehouse via API, enabling immediate ad-hoc query fulfillment.
**Business Function**: ProvideService
**Agent Checkout Support**:
- agentic-commerce-protocol
- stored-credential

## Agent Composes Headless Saas

- [Natural Language SQL Engine](/Software/Natural_Language_SQL_Engine) — AI/NLP
- [SQL Execution API](/Software/SQL_Execution_API) — Database Integration
- [Semantic Layer API](/Software/Semantic_Layer_API) — Data Modeling
- [Dataframe Processing Sandbox](/Software/Dataframe_Processing_Sandbox) — Compute
- [Interactive Charting API](/Software/Interactive_Charting_API) — Visualization
- [Messaging Webhook API](/Software/Messaging_Webhook_API) — Distribution

## Agent Sequence Diagram

```mermaid
sequenceDiagram
participant User as End User
participant BIA as Business Intelligence Agent
participant MWA as Messaging Webhook API
participant SLA as Semantic Layer API
participant NLSE as Natural Language SQL Engine
participant SEA as SQL Execution API
participant LDE as Lead Data Engineer
participant DPS as Dataframe Processing Sandbox
participant ICA as Interactive Charting API
User->>MWA: Submit ad hoc data query
MWA->>BIA: Trigger natural language request
BIA->>SLA: Fetch dbt semantic definitions
SLA-->>BIA: Return data mapping rules
BIA->>NLSE: Translate request to SQL
NLSE-->>BIA: Return generated SQL query
BIA->>SEA: Execute SQL query
SEA-->>BIA: Return raw result set
opt Anomalous data structure or semantic alteration
BIA->>LDE: Escalate for dbt semantic change approval
LDE-->>BIA: Approve and apply mapping rule
end
BIA->>DPS: Execute Pandas operations
DPS-->>BIA: Return processed dataframes
BIA->>ICA: Generate interactive charts
ICA-->>BIA: Return chart objects
BIA->>MWA: Distribute narrative briefing
MWA-->>User: Deliver business intelligence report
```

## Neighborhood

### What it consumes

- [database.do](/Software/database.do) — consumed by · Software

### What it does

- [Data Anomaly Investigation](/Tasks/Data_Anomaly_Investigation) — performs · Tasks
- [Ad Hoc SQL Analysis](/Tasks/Ad_Hoc_SQL_Analysis) — performs · Tasks
- [Metric Definition Governance](/Tasks/Metric_Definition_Governance) — performs · Tasks
- [KPI Dashboard Construction](/Tasks/KPI_Dashboard_Construction) — performs · Tasks
- [Executive Reporting Pack Assembly](/Tasks/Executive_Reporting_Pack_Assembly) — performs · Tasks
- [Translate Natural Language To SQL](/Tasks/Translate_Natural_Language_To_SQL) — performs · Tasks
- [Calculate Real-Time Margins](/Tasks/Calculate_Real-Time_Margins) — performs · Tasks
- [Execute Python Pandas Operations](/Tasks/Execute_Python_Pandas_Operations) — performs · Tasks
- [Generate Interactive Charts](/Tasks/Generate_Interactive_Charts) — performs · Tasks

### Uses tool

- [Semantic Layer Tool](/Products/Semantic_Layer_Tool) — uses tool · Products
- [Cloud Data Warehouse](/Products/Cloud_Data_Warehouse) — uses tool · Products
- [BI Visualization Platform](/Products/BI_Visualization_Platform) — uses tool · Products
- [Query Warehouse Tables](/Action/Query_Warehouse_Tables) — uses tool · Action

### Realizes

- [Self Serve Analytics Enablement](/Services/Self_Serve_Analytics_Enablement) — realizes · Services
- [Decision Support Reporting](/Services/Decision_Support_Reporting) — realizes · Services

### Optimizes

- [Query Resolution Time](/Metrics/Query_Resolution_Time) — optimizes · Metrics
- [Report First-Pass Accuracy](/Metrics/Report_First-Pass_Accuracy) — optimizes · Metrics
- [Data Request Backlog Volume](/Metrics/Data_Request_Backlog_Volume) — optimizes · Metrics

### What it uses

- [dbt](/Products/dbt) — uses · Products
- [BigQuery](/Products/BigQuery) — uses · Products
- [Slack](/Software/Slack) — uses · Software
- [Snowflake](/Software/Snowflake) — uses · Software

### Replaces this role

- [Business Intelligence Analyst](/JobTypes/Business_Intelligence_Analyst) — replaces · JobTypes

### Required skills

- [Critical Thinking](/Skills/Critical_Thinking) — requires skill · Skills
- [Data Analysis](/Skills/Data_Analysis) — requires skill · Skills
- [Database Management](/Skills/Database_Management) — requires skill · Skills
- [Information Retrieval](/Skills/Information_Retrieval) — requires skill · Skills
- [Statistical Programming](/Skills/Statistical_Programming) — requires skill · Skills

### What it addresses

- [Data Request Backlog](/Problems/Data_Request_Backlog) — addresses · Problems

### Latent gaps

- [Analytics Audit Agent](/Opportunities/Analytics_Audit_Agent) — latent gap · Opportunities
- [Data Pipeline Orchestrator](/Opportunities/Data_Pipeline_Orchestrator) — latent gap · Opportunities
- [Governance Enforcement Agent](/Opportunities/Governance_Enforcement_Agent) — latent gap · Opportunities
- [Market Intelligence Service](/Opportunities/Market_Intelligence_Service) — latent gap · Opportunities
- [Algorithmic Revenue Forecaster](/Opportunities/Algorithmic_Revenue_Forecaster) — latent gap · Opportunities
- [AI Document Extraction](/Opportunities/AI_Document_Extraction) — latent gap · Opportunities

### Composed of

- [Messaging Webhook API](/Software/Messaging_Webhook_API) — composes · Software
- [Semantic Layer API](/Software/Semantic_Layer_API) — composes · Software
- [SQL Execution API](/Software/SQL_Execution_API) — composes · Software
- [Interactive Charting API](/Software/Interactive_Charting_API) — composes · Software
- [Dataframe Processing Sandbox](/Software/Dataframe_Processing_Sandbox) — composes · Software
- [Natural Language SQL Engine](/Software/Natural_Language_SQL_Engine) — composes · Software

### Problems this exposes

- [Unstructured Data Extraction](/Problems/Unstructured_Data_Extraction) — exposes problem · Problems
- [Real-Time Revenue Forecasting](/Problems/Real-Time_Revenue_Forecasting) — exposes problem · Problems
- [Data Pipeline Fragmentation](/Problems/Data_Pipeline_Fragmentation) — exposes problem · Problems
- [Query Accuracy Validation](/Problems/Query_Accuracy_Validation) — exposes problem · Problems
- [Customer Churn Prediction](/Problems/Customer_Churn_Prediction) — exposes problem · Problems
- [Automated Data Governance](/Problems/Automated_Data_Governance) — exposes problem · Problems
- [Competitor Signal Scraping](/Problems/Competitor_Signal_Scraping) — exposes problem · Problems

### Similar Agents

- [Query Execution Engine](/Agents/Query_Execution_Engine) — similar · Agents
- [Data Engineering Agent](/Agents/Data_Engineering_Agent) — similar · Agents
- [Unified Identity Graph](/Agents/Unified_Identity_Graph) — similar · Agents
- [Event Pipeline Worker](/Agents/Event_Pipeline_Worker) — similar · Agents
- [Cross-Portfolio Query Engine](/Agents/Cross-Portfolio_Query_Engine) — similar · Agents
- [Summary Briefing Agent](/Agents/Summary_Briefing_Agent) — similar · Agents
- [Customer Service Agent](/Agents/Customer_Service_Agent) — similar · Agents
- [Billing Operations Agent](/Agents/Billing_Operations_Agent) — similar · Agents
- [Pipeline Gateway API](/Agents/Pipeline_Gateway_API) — similar · Agents
- [Unified Cost Graph Engine](/Agents/Unified_Cost_Graph_Engine) — similar · Agents

### Similar Startups

- [Adhode](/Startups/Adhode) — similar · Startups

### Similar Software

- [Google BigQuery](/Competitors/dbt_Core/Software/Google_BigQuery) — similar · Software
- [Defect Tracking Systems](/Metrics/Reliability_Analysis_Cycle_Time/Software/Defect_Tracking_Systems) — similar · Software
- [Reliability Engineering Software](/Metrics/Reliability_Analysis_Cycle_Time/Software/Reliability_Engineering_Software) — similar · Software
- [GitHub](/Competitors/Datafold/Software/GitHub) — similar · Software
- [Stripe](/Software/Stripe) — similar · Software
- [Microsoft Excel](/Software/Microsoft_Excel) — similar · Software

### Similar Metrics

- [Information Request Backlog](/Metrics/Information_Request_Backlog) — similar · Metrics
- [Time To Insight](/Metrics/Time_To_Insight) — similar · Metrics

### Similar Competitors

- [Looker](/Competitors/Looker) — similar · Competitors