Tangerine Lite v1 TAL1 · OrangePeel DAE
TANGERINE LITE V1
Deep Analysis
Deterministic Algorithmic Engine · Dual-stream computation with bottom-up synthesis, pairwise correlation, divergence detection, and convergence scoring.

Drag & drop a file here, or click to browse

Supports CSV, JSON, and TXT · Up to 50 MB · Files larger than 5,000 rows are stratified-sampled client-side

Deterministic · same file → same result, every run.

New here? Try a sample dataset below ↓

Six business datasets across five quality tiers — strong patterns down to almost-random data. Click to run the engine.

Landscaping Co.
Service Business · Tier 1

Revenue, clients, jobs completed, crew hours, regional distribution, marketing spend, satisfaction scores, and seasonal patterns.

Strong seasonal revenue pattern

E-Commerce Store
Digital Retail · Tier 1

Gross sales, returns, conversion rates, ad spend, cart abandonment, email subscribers, customer LTV, and shipping costs.

Conversion funnel + AOV dynamics

SaaS Platform
Software · Tier 2

MRR, signups, churn rate, DAU/MAU ratio, NPS scores, feature adoption, API uptime, support tickets, and enterprise deals.

Churn vs. expansion MRR

Restaurant Group
Food & Beverage · Tier 3

Covers served, food cost percentage, labor costs, review scores, delivery orders, staff turnover, waste metrics, and peak vs off-peak.

Day-of-week + weather effects

Marketing Agency
Professional Services · Tier 4

Client retainers, project revenue, billable hours, utilization rate, pitch win rate, project margins, NPS, and scope creep tracking.

Channel ROI comparison

Fitness Studio
Health & Wellness · Tier 5

Total members, class attendance, fill rates, personal training sessions, merchandise sales, retention, referrals, and trainer utilization.

Member retention + class fill rates

Deterministic

Same input always produces the same output. Every result is reproducible.

Statistically grounded

Confidence scores derived from p-values, R², and Cohen's d — not heuristics.

Algorithm-first

No LLM in the analysis loop. Pure math, fully auditable.

Understanding Your Results
Correlation Coefficient (r)
r = 1.0 Perfect positive — both metrics always move together
r = 0.7 Strong positive — high likelihood of moving together
r = 0.0 No relationship — metrics are independent
r = -0.7 Strong negative — one goes up, the other goes down
r = -1.0 Perfect negative — always move in opposite directions
Divergence Score
0.0 - 0.2 Low — categories behave similarly
0.2 - 0.4 Moderate — some meaningful differences
0.4 - 0.7 High — significant differences exist
0.7 - 1.0 Very high — categories are fundamentally different
Confidence Score
0 - 40% — Insufficient data or weak patterns
40 - 65% — Moderate evidence, patterns exist
65 - 100% — Strong evidence, significant correlations
Statistical Significance (p-value)
p < 0.05 Statistically significant — unlikely to be random
p < 0.10 Marginally significant — worth investigating
p > 0.10 Not significant — could be random chance