Package Exports
- tenzor-insights
- tenzor-insights/dist/index.js
This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (tenzor-insights) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.
Readme
๐ Tenzor Insights
Turn raw data into business intelligence โ instantly.
Tenzor Insights is an AI-powered business intelligence engine for CRM, ERP, and any structured data system.
It transforms activity logs, historical reports, and arbitrary JSON datasets into actionable insights, trend analyses, and decision-ready summaries โ powered by OpenAI, Gemini, and Claude.
โจ Features
| Feature | Description |
|---|---|
| ๐ง AI-Powered Analysis | Transforms raw logs into executive-level business insights |
| ๐ Trend Intelligence | Compares reports across time to detect business direction |
| ๐ Hybrid Intelligence | Analyzes any structured JSON data without a predefined schema |
| โ ๏ธ Risk Detection | Identifies anomalies, bottlenecks, and data quality issues |
| ๐ Growth Signals | Surfaces pipeline movement, conversions, and stagnation signals |
| ๐ค Multi-Provider | Supports OpenAI (GPT), Google Gemini, and Anthropic Claude |
| ๐งฉ Variant Modes | ADVISOR, OPERATOR, and EXECUTIVE analytical perspectives |
๐ง Modes
Tenzor Insights operates in three intelligent modes. Choose the one that fits your use case.
1. ๐ Analysis Mode
Analyze structured activity logs and generate business intelligence reports.
import { analyzeLogs } from "tenzor-insights";
const report = await analyzeLogs({
mode: "ANALYSIS",
provider: "gemini",
apiKey: process.env.API_KEY,
model: "gemini-2.0-flash",
variant: "ADVISOR", // "ADVISOR" | "OPERATOR" | "EXECUTIVE"
logs: [ /* your activity logs array */ ],
limit: 200 // optional โ default: 200
});What you get: Summary, priority scores, business impact, growth signals, risks, AI suggestions, notable users, trend indicators, and an executive insight.
2. ๐ Compare Mode
Compare multiple time-period reports to detect business trends and direction.
import { analyzeLogs } from "tenzor-insights";
const report = await analyzeLogs({
mode: "COMPARE",
provider: "gpt",
apiKey: process.env.OPENAI_API_KEY,
model: "gpt-4.1-mini",
variant: "EXECUTIVE",
reports: [
{ period: { startDate: "2026-03-01", endDate: "2026-03-07" }, report: { /* week 1 report */ } },
{ period: { startDate: "2026-03-08", endDate: "2026-03-14" }, report: { /* week 2 report */ } }
]
});What you get: Trend analysis across periods, improvements, declines, persistent issues, new risks, business trajectory verdict, confidence score, and recommendations.
3. ๐ Hybrid Mode
Analyze any structured JSON dataset โ no schema required.
Tenzor Insights will automatically infer the domain, detect entities and patterns, identify anomalies, and generate targeted business or operational insights.
import { analyzeLogs } from "tenzor-insights";
const report = await analyzeLogs({
mode: "HYBRID",
provider: "gemini",
apiKey: process.env.API_KEY,
model: "gemini-2.0-flash",
data: [ /* any JSON array or object */ ],
contextHint: "Optional: brief description of what this data represents"
});What you get: Detected domain, entity map, key insights, patterns, time-based trends, anomalies, recommendations, data quality score, and a confidence score.
How Hybrid Intelligence Works
Hybrid mode executes a 5-phase intelligence pipeline internally:
Phase 1 โ Understand the dataset (entities, fields, relationships)
Phase 2 โ Infer the domain (sales, operations, HR, finance, marketing โฆ)
Phase 3 โ Detect patterns, time trends, growth/decline, and anomalies
Phase 4 โ Generate high-quality, data-specific insights and recommendations
Phase 5 โ Assign a confidence score (0.0 โ 1.0) based on data richnessSupported domains detected automatically: sales ยท operations ยท hr ยท finance ยท marketing ยท product ยท logistics ยท general ยท unknown
๐ก
contextHintis optional but helps the engine orient faster. If the hint conflicts with the actual data structure, the engine will override it and explain the discrepancy.
๐ฆ Installation
npm install tenzor-insightsโ๏ธ Configuration
| Option | Type | Required | Description |
|---|---|---|---|
mode |
string |
โ | "ANALYSIS" | "COMPARE" | "HYBRID" |
provider |
string |
โ | "gpt" | "gemini" | "claude" |
apiKey |
string |
โ | API key for the chosen provider |
model |
string |
โ | Model name (e.g. "gpt-4.1-mini", "gemini-2.0-flash") |
variant |
string |
โ | "ADVISOR" | "OPERATOR" | "EXECUTIVE" โ default: "ADVISOR" |
logs |
array |
โ ๏ธ | Required for ANALYSIS mode |
limit |
number |
โ | Max logs to process in ANALYSIS mode โ default: 200 |
reports |
array |
โ ๏ธ | Required for COMPARE mode |
data |
any |
โ ๏ธ | Required for HYBRID mode โ any JSON array or object |
contextHint |
string |
โ | Optional domain hint for HYBRID mode |
๐ฅ Sample Input
Analysis Mode
[
{
"action": "CREATE",
"entity": "lead",
"timestamp": "2026-03-01T10:00:00Z",
"actor": { "name": "John", "role": "Sales" }
},
{
"action": "UPDATE",
"entity": "quote",
"timestamp": "2026-03-01T11:30:00Z",
"actor": { "name": "Ali", "role": "Sales" }
}
]Hybrid Mode
Pass any structured dataset โ the engine figures out the rest:
[
{ "order_id": "1001", "customer": "Jane", "total": 299.00, "status": "shipped", "created_at": "2026-03-01" },
{ "order_id": "1002", "customer": "Mark", "total": 149.50, "status": "pending", "created_at": "2026-03-02" }
]๐ค Sample Output
Analysis Mode Output
{
"summary": "Low system activity detected with only 3 actions recorded, indicating potential sales stagnation.",
"priorityScore": { "growth": 7, "risk": 2, "operations": 6 },
"keyHighlights": [
"One new lead created โ positive pipeline signal.",
"One quote updated โ progress in an existing deal."
],
"businessImpact": [
"Extremely low activity volume suggests a bottleneck in the sales pipeline."
],
"growthSignals": ["New lead created โ potential future conversion."],
"risksOrConcerns": [
{ "issue": "Critically low activity across all users.", "severity": "MEDIUM" }
],
"aiSuggestions": [
{ "suggestion": "Investigate reasons for low system engagement.", "priority": "HIGH" }
],
"ownerFocus": ["Drive consistent team utilization across all pipeline stages."],
"notableUsers": ["Ali: 2 of 3 actions โ single-user dependency risk."],
"metrics": { "actionsCount": 3, "activeUsers": 2 },
"trend": {
"activityLevel": "LOW",
"growthDirection": "STABLE",
"riskLevel": "MEDIUM",
"operationalHealth": "UNSTABLE"
},
"executiveInsight": "Current low activity levels present a direct risk to revenue growth if left unaddressed."
}Hybrid Mode Output
{
"detectedDomain": "operations",
"summary": "Dataset contains 1 order record spanning a medical e-commerce workflow. The order lifecycle is complete but abnormal re-engagement after 77 days and late-stage marketing removal events are key concerns.",
"entities": [
{
"type": "order",
"count": 1,
"keyFields": ["order_id", "order_total", "shipping_country"],
"notes": "Single order with full lifecycle tracked via nested timeline."
}
],
"keyInsights": [
{
"insight": "Order completed full lifecycle from checkout to shipment within 4 days โ efficient fulfillment.",
"category": "operational"
},
{
"insight": "Customer re-engaged 77 days post-shipment, triggering marketing removal โ churn signal.",
"category": "business"
}
],
"patternsDetected": [
{ "pattern": "Repeated member logins within short windows suggest engagement but possible confusion.", "significance": "MEDIUM" }
],
"trends": {
"hasTimestamps": true,
"timeRange": "2026-01-18 to 2026-04-07",
"direction": "STABLE",
"peakPeriod": "2026-01-20",
"notes": "Most activity concentrated in first 4 days; long gap before customer re-engagement."
},
"anomalies": [
{ "description": "ORDER_WELCOME_SMS sent with null order_id โ broken event linkage.", "severity": "HIGH", "affectedRecords": 1 }
],
"recommendations": [
{
"action": "Fix null order_id in SMS events to ensure accurate tracking.",
"priority": "HIGH",
"rationale": "Broken event linkage compromises downstream analytics and customer communication attribution."
}
],
"metrics": { "totalRecords": 1, "timelineEvents": 33 },
"dataQuality": { "score": 0.72, "issues": ["Null order_id in 2 external events", "Empty reference_id fields"] },
"confidence": 0.78
}๐ค Supported Providers
| Provider | Parameter | Example Models |
|---|---|---|
| OpenAI | "gpt" or "openai" |
gpt-4.1, gpt-4.1-mini, gpt-4o |
| Google Gemini | "gemini" |
gemini-2.0-flash, gemini-1.5-pro |
| Anthropic Claude | "claude" |
claude-3-5-sonnet, claude-3-haiku |
๐ฏ Use Cases
| Mode | Use Case |
|---|---|
| ANALYSIS | CRM log analysis, ERP activity monitoring, sales pipeline review, team performance tracking |
| COMPARE | Week-over-week trend reporting, sprint retrospectives, monthly business reviews |
| HYBRID | Order intelligence, support ticket analysis, logistics monitoring, HR data review, any custom JSON dataset |
๐ Why Tenzor Insights?
Most analytics tools show you data. Tenzor Insights tells you what it means.
| Without Tenzor Insights | With Tenzor Insights |
|---|---|
| Raw logs | Executive-ready insights |
| Numbers | Business context |
| Activity data | Growth signals & risk flags |
| Unknown JSON | Auto-detected domain intelligence |
๐ฃ๏ธ Roadmap
- ๐ฎ Predictive insights and forecasting
- ๐ Dashboard & BI tool integrations
- โก Real-time alerting
- ๐ Scheduled intelligence reports (weekly / monthly)
- ๐ Native CRM connectors (Salesforce, HubSpot)
๐ License
ISC ยฉ Greyloops