The deck title goes here

A subtitle, or the occasion · {{DECK_DATE}}

Presented by Name Surname
Gemma Analytics GmbH · internal

Section label

One claim per slide, stated as a sentence

A grouped list

  • Lead with the bold phrase. Then the explanation, which can run to a second line without hurting the rhythm.
  • Two or three items read best. Six turns the slide into a document, and the audience stops listening to you.

A figure table

LabelValue
Another labelValue
TotalSum

A statement box for the one sentence you want remembered. Keep it to a sentence.

Callout, teal for supporting context

Variants: default lavender, --teal for context, --coral for risk, --yellow for a next step.

Section label

Show the shape of the thing

Label what the diagram argues

First step What happens here, in a short clause. label the transition, not just the boxes Second step Colour can carry meaning: teal for one path, lavender for another. DASHED MEANS OUTSIDE THE SYSTEM Things referenced but not owned.

Figures

{{STAT_ONE}}Label
{{STAT_TWO}}Label
{{STAT_THREE}}Label
{{STAT_FOUR}}Label
Pass volatile figures with --var

Anything that changes between today and the presentation should be computed at build time, never typed in.

Section label

What changes in practice

Internal Sales Delivery The filled pill is the active one; outline the rest.

Before and after

WorkflowUntil nowFrom now
The activity How it went before, in muted grey How it goes now. The bold part is the change.
Another one The old way The new way. Keep both sides to one line each.
Next step

Yellow for what happens next.

Risk or caveat

Coral for what could go wrong. Say it before someone asks.

Decision paper · audience

The decision being asked for

One sentence naming the decision, so nobody has to infer it.

Option A · recommended Do it

What this option commits to, in two lines.

Option B Wait

Name the real cost of waiting, not a straw man.

Option C Stop

State this fairly too, or the panel will not trust the other two.

Timeline

Dates What happened. Past phases plain.
Dates Where we are. Lavender marks the present.
A date The gate. Coral marks a decision point.

Evidence

  • A fact that supports the recommendation.
  • Another, ideally a number you can defend.
  • What the option commits the company to.

Audience warning, if this slide is not for the main room

01

Analytics deep-dive
& project set-up

25

Your hands-on partner for data that drives real impact.

With 70+ successful projects in 5 years, our senior team brings cross-industry expertise and deep technical know-how. Fast, flexible, and always tailored.

We don't just consult, we deliver.

Our work reduces costs, boosts revenue, and improves decision-making across your company, directly impacting profits.

Founder-led. Execution-focused.

With Gemma Analytics on your side, your data function becomes a growth driver, not a cost center.

Maniko yfood Circus Group flaschenpost Wikinger Reisen
McMakler Viessmann Enpal Vaillant
28 Gemma at a glance

How we saved the project

01 Technical implementation: on-premise and cloud hybrid
02 Employee pooling and in-memory data transformations for analyses
03 Maximum automation of deployments via CI/CD pipelines
04 Overcommunication: if in doubt, take another round
05 Learning: integrate CISO, data protection and legal from the start
06 Approval by CEO
42 Solutions & learnings

How we saved the project

01 Technical implementation: on-premise and cloud hybrid
02 Employee pooling and in-memory data transformations for analyses
03 Maximum automation of deployments via CI/CD pipelines
04 Overcommunication: if in doubt, take another round
05 Learning: integrate CISO, data protection and legal from the start
06 Approval by CEO
42b Solutions & learnings

Next steps

01

Offer discussion

  • Q&A about the offer
  • Discussion of administrative details, timeline and next steps
02

Signing

Three documents are signed:

  • PoC offer
  • Data processing agreement (DPA)
  • NDA
03

Project kick-off

  • Tech kick-off: access to data sources, creation of technical resources, alignment with the technical approach
  • Business kick-off: refinement of requirements, communication channels, productivity tools
63

Module 1

Technical infrastructure

Content

  • Setting up the necessary technical infrastructure
  • Implementation of the ingestion layer
  • Setting up the transformation project and CI

Timeline & daily rate

  • {{MODULE_WEEKS}} weeks
  • Fixed price of {{MODULE_PRICE}}
66

The end of our presentation.

The start of turning data into value.

We look forward to
hearing from you!

{{CONTACT_EMAIL}}

Table of contents

Gemma at a glance 1
Analytics deep-dive 4 – 7
Case studies 8
Offer and next steps 10 – 13
23

01

Analytics deep-dive
& project set-up

Case study Auxeum: deep-dive

Phase
01

Data loading & infrastructure

02

Data modelling & transformation

03

Visualization & data validation

04

Replicate for multiple clinics

Description
  • Set up DWH, Airflow server, dlt and a connection to the clinic management tool, plus the visualization tool
  • In parallel, discussed reporting needs for phase 2 in a couple of workshops
  • Modelled patient treatment journey and patient funnels, invoice data, yearly budget targets
  • Also included data validation
  • Built operations and finance dashboards and iterated with key stakeholders on number validation
  • Replicate phases 1 to 3 for further clinics, with less effort on infrastructure and modelling because the data structure is the same
Timeline
ca. 9 days of work
ca. 9 days of work
ca. 11 days of work
Ongoing, ca. 5 days per month
31 Case study deep-dive

Tangible results for investors

Better decision-making enabled by data transparency and visualization, in particular for financials, sales and marketing data.

Efficiency gains from automated reporting and transparency, leading to EBITDA improvements.

Setting organizations up for scale through holistic data strategy and professional data pipeline setup.

Consistent approach and quality across all portfolio companies, in accordance with the investor's strategy.

Enablement of the organization through change management, training of employees and hiring support.

47 Benefits for investors

Case study Maniko

Improving scalability and flexibility by migrating from y42 to the modern data stack.

The company

Maniko is conquering the manicure and pedicure market with innovative nail products.

Problem statement

As Maniko's data and analytics needs grew, y42 increasingly struggled to keep up with the company's requirements. To ensure continued scalability and reliability, Maniko opted for a migration.

× Maniko

Gemma solutions:

Analytics infrastructure

Gemma migrated the complete analytics infrastructure from y42 to the modern data stack, in this case Fivetran, dbt Cloud, BigQuery and Tableau, so Maniko can use the best tool for each job while depending less on any one vendor.

Company-wide reporting

Gemma built on the existing analytics logic to create custom reports for each department, including management, marketing, finance, CRM, retail and operations.

Hiring support and upskilling

Gemma supported hiring Maniko's first internal data analyst, trained them, and established long-term best-practice CI/CD processes.

51 Case study

Gemma tech stack

Data creation & capturing
Data pipeline & management
Data reporting & activation

Sources

ShopifyMeta Google AdsSheets ExcelAnalytics 4 PinterestSalesforce Bing AdsKlaviyo BrazeTikTok OraclePostgreSQL Taboola

Custom databases and ERP

Loading

Apache Airflow Fivetran Airbyte

Storage

Snowflake Google BigQuery

Transformation

dbt

Reporting

LookerTableau Power BIOmni MetabaseQlik

Data consumers

Data science and AI applications

PythonForecast

Reverse ETL, for example CRM

KlaviyoBraze
Data value chain
58 Data value chain
Discovery · 29.07. to 05.08.

Agenda

01
Steering and metrics

What the minimum build carries, and who needs which figure.

02
Systems and architecture

Your target picture, our proposal, and the two decisions in it.

03
Homework on your side

What only you can define and settle.

04
Recommendation and plan

Three routes, our advice, effort and start.

Steering and metrics

The minimum build is computable

All four measures can be formed from one system. The data is there.

Net revenue and contribution

Per service line, from invoices, external costs and booked hours.

Yield per time unit

The common currency across every billing model, total and per line.

Productive hours ratio

Per line, from time bookings, absences and working-time models.

Hours variance

Actual hours against the historical reference for the same order type.

Three definitions still need sharpening: “service line”, real cost rates instead of industry rates, and the denominator of the productive hours ratio. That is detail work on terms, not a missing data source.

Forward-looking
capacity

All four measures look backwards. What is needed is free capacity per line, two to three months ahead — in days and euros, not as a ratio. No system holds it today, and it needs a decision from management, not technology.

Systems and architecture

Your target picture and our proposal

Architecture v8
Sources
Time tracking
ERP
Accounting
HR, surveys
Semantic layer
Dataflows in the reporting tool
Reporting
Dashboards

No warehouse, no orchestration, no version control. The metric logic lives inside the reporting tool.

Our proposal
Sources
Time tracking
ERPlater
Accountinglater
HR, surveyslater
Load and orchestration
Airflow
Data warehouse
SnowflakePostgreSQL as the alternative
Semantic layer
dbt
Reporting
Lightdash
Existing toolstays possible

Transformations and metric definitions live versioned in Git: traceable, testable and independent of the reporting tool.

Systems and architecture · decision 1 of 2

Warehouse

Three workable routes. We recommend one and say what the other two save.

Our recommendation
Snowflake
per month, estimatedca. 300 €
  • access and role management
  • infrastructure as code
  • backup and retention included
  • grows with further use cases
  • data protection assessment needed
Alternative in the cloud
PostgreSQL, managed
per month, estimated60 to 120 €
  • cheaper than the recommendation
  • backup and retention included
  • little role and access management
  • a ceiling at more data and complexity
  • data protection assessment needed
Alternative on site
PostgreSQL, on-prem
per month, external0 €
  • data never leaves the building
  • no licence or cloud cost
  • operation and backup fall to you
  • little role and access management
  • a ceiling at more data and complexity
Homework on your side

What has to be settled on your side

The problem is not the dashboard, it is the processes in front of it. Only you can settle these.

Task Blocks Owner
Technically solvable
A read API on the time tracking system every data access Platform team
To be decided by the business
Real cost rates per qualification group contribution and yield per time unit Commercial
Define service lines and cost centres all four measures Platform, commercial
Classify projects as internal or external the productive hours ratio Platform team
Introduce forward booking of capacity the capacity view Platform, management
Recommendation and plan

Three options for the minimum build

We implementone-off, no standing contract {{PRICE_BUILD}}
  • Covers the whole stack: connection, orchestration, warehouse, models, reporting
  • Your developer is in from day one, their upskilling is in the price
  • You run and extend it afterwards, without us
We write the build planyou implement {{PRICE_PLAN}}
  • Target architecture, tool choice and build order as a manual
  • Implementation, testing and operation stay entirely with you
  • Subsidised under the same programme as this project
You do it aloneno further consulting
  • This report is the basis
  • The learning curve and your own resource plan stay open
Our recommendation

We implement, your developer builds alongside and takes it over. It is the route that leaves capability behind rather than dependency.

{{PRICE_BUILD}}fixed price
{{WEEKS}}build
{{START}}possible start