AskTable Built-in Skills and Agents: Out-of-the-Box Data Analysis Capabilities

Do you have these troubles:

• •You have data but don't know where to start analysis
• •Every report takes a lot of time to organize data and write
• •Anomalies happen but you're always late, missing the best adjustment window
• •Want to use AI for data analysis but don't know how to tune it

The root cause of these problems: General AI doesn't understand your business, let alone have the professional analysis capabilities you need.

AskTable's solution: Transform professional data analyst capabilities into reusable skills and agents.

---

I. AskTable's Skill System

What are Skills?

Skills are AskTable's built-in professional analysis capabilities, like an experienced analyst on standby. You don't need to tell it "how to analyze", just tell it "what to analyze".

11 Built-in Skills

1. Anomaly Detection - Discover Data Anomalies at First Moment

Automatically identify abnormal fluctuations in data, no longer relying on manual inspection.

Typical scenarios:

• •"Why did today's sales suddenly drop 20%?"
• •"Is this month's conversion rate abnormal?"

What it can do:

• •Calculate historical baseline, automatically identify metrics deviating from normal range
• •Mark anomaly time points, give deviation percentage
• •Analyze possible business reasons, recommend drill-down dimensions

---

2. Prediction Trend - Seeing the Future

Predict trends based on historical data to assist planning decisions.

Typical scenarios:

• •"What's next month's sales approximately?"
• •"Can we complete the quarterly target?"

What it can do:

• •Identify trends, seasonality, cycle patterns
• •Give point predictions and confidence intervals
• •Warn about interference from anomalous values on predictions

---

3. Drill-Down Metrics - Locating Problem Root Causes

Decompose from summary metrics layer by layer to find specific causes of problems.

Typical scenarios:

• •"Sales dropped, exactly which area is dragging it down?"
• •"Which region, category, time period has the problem?"

What it can do:

• •Multi-dimensional decomposition by time/region/category/person
• •Find dimensions with largest contribution difference
• •Progress layer by layer until finding specific actionable problem points

---

4. Comparative Analysis - Finding Gaps and Opportunities

Horizontal comparison to discover differences and opportunities.

Typical scenarios:

• •"Where are the gaps between us and competitors?"
• •"Why does Store A perform better than Store B?"

What it can do:

• •Time comparison (QoQ/YoY/fixed base)
• •Spatial comparison (region/store/department ranking)
• •Plan comparison (actual vs target/budget)
• •Identify key drivers of differences

---

5. Attribution Analysis - Quantifying Each Factor's Impact

Analyze core drivers of metric changes and quantify each factor's contribution.

Typical scenarios:

• •"This month's profit increased, exactly which factors contributed?"
• •"Revenue growth due to price increase or volume increase?"

What it can do:

• •Decompose each factor's incremental contribution
• •Calculate each factor's elasticity coefficient
• •Identify internal/external/occasional factor impacts

---

6. Stress Testing - Evaluating Extreme Scenarios

Simulate different hypothetical scenarios to evaluate business performance under extreme conditions.

Typical scenarios:

• •"If revenue drops 30%, how long can we survive?"
• •"What's the impact of raw material price increase 20% on profit?"

What it can do:

• •Design optimistic/baseline/pessimistic three scenarios
• •Find most critical impact factors
• •Evaluate risk exposure and give response plans

---

7. Cycle Analysis - Understanding Fluctuation Patterns

Identify cycle patterns in data to understand the logic behind fluctuations.

Typical scenarios:

• •"Why does sales drop every weekend?"
• •"Does our business have obvious seasonality?"

What it can do:

• •Identify short/medium/long-term cycles
• •Calculate seasonality indices
• •Trend analysis after removing seasonality

---

8. Report Orchestration - Auto-Generate Professional Reports

Organize data analysis results into structured reports.

Typical scenarios:

• •"Generate this week's sales report for me"
• •"Make a monthly operations analysis presentation"

What it can do:

• •Generate structured reports with executive summary + key metrics + anomaly analysis + suggestions
• •Automatically include QoQ/YoY/target comparisons
• •Conclusions first, data follows

---

9. Metric Interpretation - Explaining Data in Plain Language

Translate technical metrics into language business people can understand.

Typical scenarios:

• •"Is this 3.2% conversion rate good or not?"
• •"Help me explain what this data means"

What it can do:

• •"Conversion rate 3.2%" → "3.2 out of every 100 visitors complete a purchase"
• •Verify if data is within reasonable range
• •Supplement business context behind numbers

---

10. Data Quality Detection - Ensuring Analysis Reliability

Proactively detect data quality to ensure analysis results are accurate.

Typical scenarios:

• •"Is this data reliable?"
• •"Are there problems with recent data?"

What it can do:

• •Detect null values, missing data ratio
• •Identify data definition inconsistency issues
• •Give data quality scores

---

11. Business Language Generation - Telling a Compelling Data Story

Transform data into warm business narratives.

Typical scenarios:

• •"Write me a paragraph to report to the boss"
• •"Make a compelling data presentation"

What it can do:

• •Conclusions first, data follows
• •Use analogies to explain complex metrics
• •Use comparisons to enhance data impact

---

II. 9 Industry Agents

What are Agents?

Agents are skill combinations plus role settings and business knowledge, forming a "knowledgeable" virtual colleague. You don't need to configure skills one by one, just select an appropriate agent and it can start working.

Agents Overview

---

Retail Operations Analyst

Applicable scenario: Daily retail store operations analysis

You are an experienced retail operations advisor.

When you start, you proactively help:
- Integrate multi-source data like sales, foot traffic, and inventory
- Identify high-growth/low-growth stores
- Discover anomalous time periods and categories
- Provide actionable improvement suggestions
- Auto-generate daily/weekly/monthly reports

One sentence: Transform regional manager experience into an always-available digital advisor.

---

E-commerce Data Monitor

Applicable scenario: Daily e-commerce operations monitoring

You are a diligent e-commerce data monitoring assistant.

When you start, you proactively help:
- Real-time monitoring of sales, traffic, conversion, average order value
- Compare your data with market benchmarks
- Immediately alert when metrics exceed thresholds
- Auto-generate daily data reports
- Special monitoring for important nodes (promotions/events)

One sentence: Your 7×24 operations partner who never takes a break.

---

Financial Data Analyst

Applicable scenario: Enterprise financial analysis

You are a professional financial analyst.

When you start, you proactively help:
- Accurately interpret three statements (balance sheet, income statement, cash flow statement)
- Calculate key financial ratios
- Alert solvency risks, liquidity risks
- Track budget execution progress
- Break-even analysis and ROI evaluation

One sentence: Transform CFO-level analysis capability into reproducible standard service.

---

Executive Data Assistant

Applicable scenario: Executives' one-click operations data summary

You are a concise and efficient executive data assistant.

When you start, you proactively help:
- Core operational metrics overview, compared with target/same period
- Conclusions first, at most 3 key findings
- Answer any data questions anytime
- Daily operations morning report, weekly operations report
- Immediate alerts for major anomalies

One sentence: Let executives control the whole picture spending just 5 minutes daily.

---

Traffic Light Analyst

Applicable scenario: Business line health tracking

You are a sharp business health advisor.

When you start, you proactively help:
- Set traffic light thresholds for each metric together with you
- Check each metric's status automatically every day
- Proactively alert and analyze causes when red/yellow lights appear
- Provide improvement suggestions with priority ranking

One sentence: Give every business leader a health advisor.

---

III. How to Use These Capabilities and Templates

Method 1: Directly Use Built-in Agents

If you see a suitable agent template, just use it. AskTable will automatically load the corresponding skill combination. You only need to:

1. •Select agent template
2. •Connect your data sources
3. •Agent starts working

Method 2: From Agent to Skills, Adjust as Needed

If built-in agents don't fully match, you can:

1. •Select the closest agent as starting point
2. •Add or remove skills as needed
3. •Adjust role settings and focus metrics
4. •Save as "My Agent"

Method 3: Learn from Conversations, Build Exclusive Skills

During use, you can anytime say:

• •"When analyzing sales in the future, remember to include QoQ and YoY"
• •"I focus on East China region data"

AskTable will quietly remember these preferences and form your exclusive skills.

---

IV. Customer Cases

A Certain Chain Retail Brand

Pain point: 200 stores, daily data aggregation takes 2 hours, anomaly discovery always lags.

Solution: Deploy Retail Operations Analyst + Traffic Light Analyst.

Effects:

• •Daily report generation: 2 hours → 5 minutes
• •Anomaly discovery: From 1 day lag → Real-time alerts
• •Regional managers save 1.5 hours daily on data processing

A Certain E-commerce Company

Pain point: Operations team monitors the market daily, misses best adjustment windows while watching.

Solution: Deploy E-commerce Data Monitor + Anomaly Detection Skill.

Effects:

• •Core metrics real-time monitoring, anomalies alerted within 5 minutes
• •Daily auto-generated data reports
• •Operations transformed from "monitoring tool" to "strategy optimizer"

---

V. Quick Start

1. •

   Try agents: In AskTable console, select "Create Agent", browse built-in templates, find the closest to your needs

2. •

   Connect data: Select data sources you need to analyze (database, data warehouse, Excel, etc.)

3. •

   Start working: Agent begins analyzing, proactively pushes daily reports/alerts to you, or you ask questions anytime

4. •

   Adjust as needed: If something's not right, adjust skill combinations and role settings anytime

---

AskTable's goal isn't replacing data analysts, but letting everyone have data analysts' professional capabilities.

Built-in skills and agents are the first step toward this goal.