Gemma Beyond Data
From data consultancy to senior technology partner — a founder's thesis on where Gemma needs to go, and how we get there.
AI is compressing pure data delivery. DWH setup is faster and cheaper every quarter. Meanwhile, clients already pull us beyond data — process automation, bespoke tooling, AI applications. We do some of this reactively. The market is growing fast: Heuristiq, Lucid Labs, Nexamind — all focused narrowly on "we build AI agents for you." We can offer something they can't. This page lays out what and how.
Conviction · Direction
Where is Gemma going?
Data & AI hybrid
Data-rooted transformation
Full-service consultancy
Data is our origin, not our boundary. We solve real business problems end-to-end — with the right technology.
Pure data work is commoditising fast. New competitors focus narrowly on building AI agents — Heuristiq, Lucid Labs, Nexamind all say "we bring you AI." But real value isn't just agents. It's understanding the business, making the data work, choosing the right tool, and shipping what lasts. That's a bigger market — and our 6 years of data DNA is a structural advantage no 18-month-old AI studio has.
Decision · Rollout
How do we take this to market?
Internal only for now
Phased Q2 → Q4
Full rebrand now
Q2–Q3: Align team, enable client owners, run pilots. Q4: External repositioning backed by real evidence.
The direction is set — this is about pace and communication. Everyone at Gemma needs to understand where we're heading. Moving too fast externally without delivery proof = empty rebrand. The phased approach: equip the team now, build conviction through real client wins in Q2–Q3, then reposition externally in Q4 from a place of strength.
The Claim — draft
Gemma makes digitalization real.
We understand your business, find where technology creates real value, and build it — data, AI, automation, bespoke software. We think with you, then we ship for you.
Klartext
Honest about what works and what doesn't. We recommend the right solution — including when it's not AI.
Prototype to Production
Fast to show what's possible. Serious about making it last. We ship and we stay.
The Full Stack
Data, AI, process redesign, bespoke software — we build every layer your problem needs. Not just agents.
Three Buyers We Build For — sales angle adjusts per persona
Persona 01
Mittelständler mit KI-Mandat
The board says "do AI" — that's why they call. What they actually need: an honest assessment, a data foundation, and real first use cases that deliver ROI, not slideware.
Door: "AI". Delivers: AI-readiness, DWH, governance, first working use cases.
Persona 02
Scaling Start-up / Scale-up
Past spreadsheets, no data or tech team. Needs a senior technology partner who handles data, dashboards, and the tools in between.
Sells on: "We are your technology partner — from first dashboard to production AI."
Persona 03
The Operational Leader
Runs ops, finance, or logistics. Has real process pain — manual work, broken tools, duct-taped workflows. Doesn't want AI hype. Wants someone who understands the problem first.
Sells on Klartext: "We pitch solutions, not technology brands."
What this is not: Data remains our craft and our largest revenue stream. This is not about abandoning analytics engineering — it's about removing the artificial boundary that limits what we do for clients. Every current analytics engineer has a home in this vision.
Evidence, Plan & Implications
What's already happening, the rollout timeline, and what this means for team and hiring.
This Is Already Happening — "beyond data" work, most of it unplanned
Existing Clients
Tasty Urban
Credit note process automation, ops tooling prototype.
Infoniqa
Streamlit app on DWH for sales workflows.
Natsana
Lovable-built sales dashboard, beyond classic BI.
Terras Tiefbau
We did the data — but Commasoft built the AI Streamlit app. Why not us?
Leads & Pipeline
Hellomed
Multi-agent AI system for process automation. Deep in client ops.
Deutsche InfraSoft
Wants to use AI internally — but has no idea where to start or how.
ROTOP Pharmaka
"Give everyone Claude so they can do data analysis." Needs guidance on what that actually means.
None of this was strategically planned. Clients pull, leads ask, and in one case (Terras) we left money on the table by not offering what we could have. Internally, Gemmbot and Gemma OS prove we can build bespoke tooling. Imagine what happens when we actively sell this across every client.
Q2 · Now → June
Align & Enable
eMR alignment (this meeting). Communication to full team. Launch upsell initiative with B2B sales coach. Every client owner reviews full portfolio for "beyond data" opportunities.
Q3 · Jul → Sep
Sell & Ship
Work with B2B sales coach to pitch "beyond data" to clients. Every client owner: at least one sold & shipped pilot. Document what resonates, what we learn about delivery.
Q4 · Oct → Dec
Reposition Externally
Updated website, pitch deck, case studies. External messaging reflects the new positioning — backed by real delivery evidence from Q3 pilots.
We do
- DWH & data modeling
- AI-readiness & governance
- Pipelines & data integration
- Process analysis & digitalization
- Bespoke internal tooling
- AI applications & automation
- MLOps & model deployment
- Consulting attached to delivery
We don't do
- CRM / ERP implementation (→ partner)
- Tracking & marketing tech (→ partner)
- Performance marketing
- Strategy consulting without delivery
- Body-leasing / junior staffing
- Legacy infra for non-clients
- Customer-facing prod software — open
- Standalone AI strategy workshops — open
Team Skills
Client owners learn to sell beyond data. Engineers build prototyping + AI/automation skills. SCOs need to go deeper into client processes.
Hiring Profile
Next hires: senior "builders" who can prototype and ship across data + AI + software. Less pure AE, more full-stack delivery.
Client Strategy
Account reviews shift: "what problems can we solve?" not just "what data work is next?" Upsell playbook becomes standard practice.
Revenue Model
Broader scope → larger average deal size. Mix may shift from pure retainer toward project + retainer combinations.
Brand & Messaging
Q4 external repositioning. New website, pitch deck, case studies. Internal language shifts now — "technology partner," not just "data consultancy."
Internal Tooling
Gemma OS + Gemmbot as proof-of-concept and capability demonstrator. Dogfood what we sell. Build internal best practices for AI/automation delivery.
Discussion Agenda for eMR
- Does this direction resonate? What worries you about it?
- Look at your full client portfolio: where do you see "beyond data" opportunities? Name them.
- What would you need to sell this to a client? What's missing today?
- Does this timeline feel right? Too fast? Too slow?
- How does this align with initiatives we're already running? What needs to change?
- How and when do we communicate this to the rest of Gemma?
- Where do we draw the line — what should Gemma never do?
- What capability are we missing? What's the first hire?