EVA app icon
EVA / WORKSPACEKISSA SAMPLE
PUBLIC SAMPLE · PLAN SWITCH IS AN INTERFACE PREVIEW, NOT A SUBSCRIPTIONKISSA COFFEE HOUSE / 喫茶こもれびbrand context from supplied Kissa materials · analytics synthetic / anonymized
01 / START WITH THE MESS

RAW DATA.
BEFORE THE DASHBOARD.

Most businesses do not start with a clean BI model. They start with exports, spreadsheets, duplicated labels, inconsistent dates and missing fields. This KISSA sample intentionally keeps that mess visible. Product names and brand context follow the supplied Kissa Komorebi materials; transaction and performance records are generated for demonstration only.

ROWS570
FIELDS32
PRODUCT FAMILIES4
INTENTIONAL DUPLICATES9
＋
Try your own CSVOpen EVA DataLab · small-data demo
 Download messy KISSA CSV ↓
DATA HEALTH / BEFORE EVAKISSA_RAW_MESSY_SAMPLE.CSV
5 SOURCE TYPES
brand-aligned products5 source formats9 duplicate copiesmixed dates + JPY strings
transaction_dateorder_idproduct_nameqtynet_salessales_channelsource_file
Sep 02 2026KK-05000Seasonal Bake1JPY 450CounterPOS_JULY.csv
18/07/2026KK-00582Café Latté¥530IN_STOREmanual_sheet
08-19-2026KK-05001House Blend Dark Roast1JPY 1,450TAKE_AWAYpos-export.csv
Jul 18 2026KK-05002Drip Coffee1350take awaydelivery_weekly.xlsx
22/08/2026KK-05005Vanilla LatteMOBILE ORDERmobile_orders.JSON
SCHEMAMISSINGNESSDUPLICATESALIASESTYPES
02 / EVA DATA

MAKE THE SOURCE
DECISION-READY.

Scanning schema and source structure…

RAW47quality flags
01PROFILE

Infer structure, missingness, duplicate rows and type inconsistencies.

02STANDARDISE

Normalise dates, products, channel names and numeric fields.

03VALIDATE

Run QA checks and build a reporting-ready business layer.

04MODEL

Create reliable metrics EVA AI and EVA LABS can use.

When the sample is ready, EVA moves to the next step automatically.

03 / EVA AI

ASK FOR THE
BUSINESS ANSWER.

In the full product you can type a question. In this public sample, choose what you want EVA to build from the cleaned KISSA data.

EVA AICHOOSE A QUESTION
DEMO
OUTPUTChoose a question.

EVA will select the workspace modules that answer it.

04 / EVA LABS / KISSA SAMPLE

DECISION WORKSPACE.

SIMULATED PLAN VIEW · NOT A PAID ACCOUNT
NET REVENUE
¥1.84M

Modeled sample period · +12.4%

ORDERS
1,284

43 / day average

AVG. TICKET
¥1,433

Blended coffee + food + goods

REPEAT GUESTS
42%

Loyalty / repeat signal

REVENUE / PRODUCT
CAFÉ LATTE¥425K
DRIP COFFEE¥361K
VANILLA LATTE¥296K
BEANS + GOODS¥240K
BRAND / BUSINESS CONTEXT
Kissa Komorebi brand
A QUIET CUP, A GENTLE MOMENT.

Komorebi means sunlight filtering gently through leaves. EVA keeps the supplied brand promise, menu structure and merchandise context visible beside the synthetic operating data.

COFFEE · FOOD · BEANS · GOODS
EVA AI / PRIORITY

Protect the quiet experience before adding more demand.

The synthetic sample flags peak service time and ingredient coverage as the immediate constraints. Growth only helps if the brand experience stays calm.

FIX FIRSTCAPACITY + STOCK
DATA HEALTH / SOURCE CONTROL
570synthetic sample rows
9intentional duplicate copies
4product families modeled
5source formats reconciled
READINESSREPORTING-READY AFTER EVA DATA
SCHEMA
32

fields across transaction, product, customer, stock and source metadata.

QUALITY BOUNDARY

The public demo does not claim causality. Descriptive findings, held-out model evidence and management recommendations are kept visibly separate.

HOURLY SALES / SAMPLE DAY
08101214161820
CUSTOMER SEGMENTS

Repeat / loyalty42%

Occasional33%

New / unknown25%

MODEL LAB / FUTURE SCENARIO
PROJECTED REVENUE¥5.55M
PROJECTED GROSS PROFIT¥3.55M
GROSS MARGIN64%

Protect peak-period service quality first, then test a simple coffee + seasonal bake bundle.

Illustrative gross profit = projected revenue minus estimated cost of goods sold (COGS). This does not include salaries, rent, taxes, other operating expenses or financing costs, and is not net profit. Synthetic data only; not a validated forecast.

DECISION BRIEF / NEXT ACTIONS
  1. 01
    Protect peak service capacity and ingredient coverage.

    Reduce lost demand without making the calm Kissa experience feel rushed.

  2. 02
    Test one brand-safe coffee + seasonal bake bundle.

    Measure ticket lift and contribution margin while keeping the offer simple.

  3. 03
    Re-measure after 30 days.

    Compare demand, stockouts, repeat visits and margin against the baseline.

MANAGEMENT NOTES

01Check peak service and ingredient coverage before next campaign.

02Review coffee + seasonal bake bundle margin after first test.

AI + HUMAN / BUSINESS

Turn the signal into a management move.

EVALUATION can review the data, explain commercial implications and convert EVA output into a marketing, pricing or operations action brief.

Ask for human insight ↗
ASK EVA AI

Ask about revenue, what to fix first, future risk or a management summary.