DataAnalyzer

Category: Analysis | Module: mycontext.templates.free.analysis

Systematically analyzes data and extracts actionable insights. Supports 5 intents (executive, analyst, operations, summary, comprehensive) and 3 investment levels (quick, standard, thorough) for fine-grained control over what sections are produced and how deep the analysis goes. Provides descriptive statistics, trend detection, anomaly identification, correlation analysis, comparative segmentation, and prioritized recommendations — always with explicit evidence and confidence ratings.

When to Use

• Business metrics analysis (revenue, engagement, retention)
• Product analytics review
• Sales data investigation
• Operations and incident triage
• Log data pattern analysis
• Any structured dataset where you need to go from numbers to decisions

Quick Start

from mycontext.templates.free.analysis import DataAnalyzer

analyzer = DataAnalyzer()

# Full comprehensive analysis (default)
ctx = analyzer.build_context(
    data_description="Monthly sales data for past 12 months: Jan $120k, Feb $95k, Mar $140k, Apr $155k, May $148k, Jun $162k, Jul $108k, Aug $115k, Sep $178k, Oct $195k, Nov $210k, Dec $185k",
    goal="Identify growth patterns and seasonal trends",
)
result = ctx.execute(provider="openai")
print(result.response)

# Executive briefing — focused, fewer tokens
result = analyzer.execute(
    provider="openai",
    data_description="Monthly sales data...",
    goal="Key takeaways for leadership",
    intent="executive",
    investment="quick",
)

Intents

The intent parameter controls which sections appear in the output:

Intent	Sections	Best For
executive	4	C-level briefings — overview, insights, viz, recommendations
analyst	4	Data science — overview, statistics, patterns, correlations
operations	3	Ops triage — overview, anomaly detection, recommendations
summary	3	Quick decisions — overview, insights, recommendations
comprehensive	11	Full audit — all sections (default)

Investment Levels

The investment parameter controls how deep each section goes:

Level	Token Budget	Behavior
quick	~1,500	Concise bullet points, max 3 insights
standard	~3,000	Default depth, full section coverage
thorough	~5,000	Exhaustive analysis with detailed evidence chains

Intent and investment compose freely — any of the 15 combinations is valid.

Methods

build_context(data_description, goal, context, intent, investment)

Parameters:

Parameter	Type	Default	Description
data_description	str	""	Description of the data — raw numbers, summaries, or structured text
goal	str	"Extract insights"	What you're trying to learn from the data
context	str | None	None	Additional context: data source, time period, known events
intent	str	"comprehensive"	Report type: executive · analyst · operations · summary · comprehensive
investment	str	"standard"	Depth level: quick · standard · thorough

execute(provider, data_description, goal, context, intent, investment, **kwargs)

result = analyzer.execute(
    provider="openai",
    data_description="""
    Q1: 1200 signups, 340 conversions (28%)
    Q2: 1500 signups, 375 conversions (25%)
    Q3: 1800 signups, 360 conversions (20%)
    Q4: 2100 signups, 315 conversions (15%)
    """,
    goal="Understand conversion rate decline despite signup growth",
    context="SaaS product, B2B, annual contract",
    intent="analyst",
    investment="thorough",
)

Analysis Framework

The 11-section analysis covers:

1. Data Overview — Type, coverage, variables, quality assessment
2. Descriptive Statistics — Central tendency, dispersion, distribution shape
3. Pattern Detection — Trends (direction, magnitude, timeframe), seasonality cycles, clusters
4. Anomaly Detection — Outliers with context, severity, and possible causes
5. Correlation Analysis — Variable relationships with causation-vs-correlation distinction
6. Comparative Analysis — Segment comparison table with key differences
7. Key Insights — 3–5 ranked insights, each with evidence, confidence rating, and recommended action
8. Hypotheses — Explanations for detected patterns with tests
9. Data Limitations — Gaps, biases, and what cannot be concluded
10. Recommendations — Immediate actions + further investigation
11. Visualization Suggestions — Chart types optimized for each finding

Confidence Ratings

Every insight includes an explicit confidence level:

Confidence	Meaning
High	Pattern is clear, data is sufficient, alternative explanations are unlikely
Medium	Pattern visible but data limitations exist or alternatives are plausible
Low	Suggestive but insufficient to conclude — needs more data

Examples

Executive Briefing (quick)

result = analyzer.execute(
    provider="openai",
    data_description="""
    Monthly Active Users by segment over 6 months:
    Enterprise: 450, 462, 471, 480, 488, 494 
    SMB: 2100, 1980, 1870, 1750, 1640, 1530
    Startup: 890, 920, 960, 1010, 1080, 1160
    """,
    goal="Segment health for board meeting",
    intent="executive",
    investment="quick",
)

Analyst Deep-Dive (thorough)

result = analyzer.execute(
    provider="openai",
    data_description="Product category revenue data with 24 months history, customer acquisition cost by channel, and return rates by product line",
    goal="Identify highest ROI product categories and channels for budget allocation",
    context="Fashion e-commerce, $5M annual revenue, US market",
    intent="analyst",
    investment="thorough",
)

Operations Triage

result = analyzer.execute(
    provider="openai",
    data_description="Server error rates past 24h: 0.1%, 0.2%, 0.1%, 0.8%, 2.1%, 4.5%, 3.2%",
    goal="Identify anomalies and immediate remediation steps",
    intent="operations",
    investment="standard",
)

Generic Prompt Mode

prompt = analyzer.generic_prompt(
    data_description="Monthly churn rates: 2.1%, 2.3%, 2.0%, 2.8%, 3.2%, 3.1%",
    goal="Identify if churn is trending up",
)

Output Format

Insights are structured for immediate action:

**Insight #1**: Conversion rate is declining despite signup growth
- Evidence: Q1 28% → Q4 15%, consistent 3-5% quarterly decline
- Confidence: High
- Significance: Revenue impact is 40% below what signup growth implies
- Action: Audit onboarding flow for friction introduced in Q2 redesign

**Insight #2**: SMB segment is churning systematically
- Evidence: 27% decline in MAU over 6 months (-95 users/month average)
- Confidence: High
- Significance: Represents $180k ARR at risk annually
- Action: Conduct exit interviews with churned SMB accounts

API Reference

Method	Returns	Description
build_context(data_description, goal, context, intent, investment)	Context	Assembled context with parameterized sections
execute(provider, data_description, goal, context, intent, investment, **kwargs)	ProviderResponse	Execute analysis with auto token budget
generic_prompt(data_description, context_section, goal)	str	Zero-cost prompt string

Class Attribute	Type	Description
INTENTS	dict	Maps intent names → section lists
INVESTMENTS	dict	Maps investment levels → constraints + token hints
SECTION_BLOCKS	dict	All 11 section templates