ALEX
SEIPEL
I work out where AI pays off.
Then I build it.
Ten years of engineering, from a two-person startup to a bank's core platform. Hire me for a week of advice, or for the whole build.
10+
Years of experience
06
Industries
04
Years as co-owner & CTO
01ServicesConsulting and delivery
01
AI consulting
AI moves fast, and most of it will not matter to you. My job is to work out which parts do, and to say so plainly.
Assessment
I look at how your team actually works and find the tasks where a model pays off. Sometimes the honest answer is that none of them do.
Tool selection
Buy a tool, combine several, or build your own. I tell you which one fits, what it costs to run, and what you are committing to.
Training
A tool nobody uses is just a subscription. I run the sessions, write the internal guidelines, and stay reachable afterwards — it is what I already do inside a bank.
02
AI software
Off-the-shelf tools cover the easy cases. When your problem is specific, I design and build the software for it.
Better interface
A chat box is a weak interface for work you already understand. A purpose-built screen is faster and leaves less room for error.
Model selection
Model prices differ by an order of magnitude, and so does the quality. I benchmark them on your data and use the expensive one only where it pays.
Data protection
The model gets the least access that still works, and the GDPR side is solved, not waved away. If your data cannot leave the building, I run an open model on your own hardware.
03
Software development
Not every problem needs a model. I also build ordinary software: web applications, APIs, online shops, internal tools.
Prototype first
You get something you can click on early, while changing direction is still cheap. Then we harden it into a product your team can maintain.
Development with AI
Agents write most of the code now — I work with Claude Code and its peers daily. The work moved to the specification and the review, and that part is still mine.
Architecture
AWS, Terraform, Docker, CI/CD. I once had two weeks to make a platform ready for a national TV audience — auto-scaling, WAF, failover included.
E-commerce
Headless Shopify, subscription models, Stripe, and the marketing stack around them. I ran a shop's entire tech for four and a half years.
Accessibility
The BFSG made WCAG conformance law in 2025. I build accessible interfaces at banking standards, and I audit and fix existing ones.
02Experience2016 to now
I build software that has to hold up in production.
Ten years, from a two-person startup to a bank. Four and a half of them as co-owner and CTO at Brandl Nutrition. I owned the whole stack there: the engine that turned a questionnaire into a personal recipe, the shop, the AWS infrastructure, the order automation.
Then a year of consultancy — supply chains, a certification group's design system, digital healthcare — and today a core banking frontend at Commerzbank that millions of customers use. Before all that I trained neural networks and shipped ML pipelines at Immowelt, part of the Axel Springer group. The AI side is not new ground for me.
Off duty I lift weights and plan my meals in a spreadsheet, down to the gram — it is how I ended up at a nutrition startup in the first place. Measure first: that is also how I build software.
- LocationPrague, Czechia · CET
- RemoteEurope
- NowCommerzbank AG · Engineer
- BeforeCo-owner & CTO · Brandl Nutrition
- LanguagesGerman · English
- Emailalex@seipel.me
- Transaction Overview is a core component of the online banking, used by millions of customers. I maintain and extend its React frontend, and moved its toolchain from React Scripts to Vite.
- For a new advisory platform with AI features I architected the React frontend and built several modules end to end on my own.
- Accessibility at banking standards: WCAG conformance is part of every feature I ship.
- I help drive the adoption of AI tooling in engineering and train colleagues on it. I also onboard new developers and write the OpenAPI specifications.
- Seedtrace: a data import tool for supply chain data — React, AgGrid, GraphQL. My first direct client work: I ran the requirements workshops and defined the features before building them.
- TÜV Nord: a token-based design system, end to end — Figma tokens synced via Git, StencilJS web components, wrapper packages for React, Angular and Vue.
- Neurocare: a therapy platform for clinics worldwide — I learned Angular on the job, was productive within days, and mentored a junior developer.
- I built the engine that turned a customer questionnaire into a personal supplement recipe — and automated fulfilment from checkout to printed label, production and DHL hand-off included, with a Django backend.
- A shot at the TV show “Die Höhle der Löwen” meant being ready for a national traffic spike on two weeks’ notice. I moved everything from one rented server to auto-scaling AWS, managed with Terraform.
- I replaced Magento with headless Shopify behind a Gatsby frontend, and ran the whole marketing stack — analytics, A/B tests, newsletter automation.
- As co-owner I shaped the product too: the Pure Protein I pushed for became the best seller.
- Fraud prevention for one of Germany’s biggest property portals (Axel Springer group): a mobile TAN flow and identity-verification integrations.
- “Markt und Preise”: a public analytics tool where users compose their own charts over Germany-wide rent and purchase prices — C#/.NET on ClickHouse.
- ML end to end, from data to deployment: price estimation, image classification, data extraction from energy certificates — TensorFlow, Keras, PyTorch, scikit-learn.
- The data side: SQL Server, Vertica, Couchbase, MongoDB, ClickHouse — plus a real-time processing library we built for MongoDB.
- My first commercial position. I worked on search engine optimization and on content.
- I built a search prototype with Apache Solr and learning-to-rank algorithms.
- I examined Twitter log data with R for information retrieval experiments.
On the side
impflotse.de ↗
Appointment booking for vaccinations at pharmacies. I build the platform as the sole engineer — AI-assisted from the first commit, with GDPR done properly: legal logging, automated data disclosure.
Affiliate SEO
Earlier: an affiliate site that ranked, earned, and taught me search engine optimization hands-on.
From my employment references
“Mr Seipel’s performance earned our full recognition at all times and in every respect.”
“He found excellent solutions for all arising problems without exception.”
03StackWhat I build with
AI 01
- LLM APIs
- Claude
- GPT
- Agents
- MCP
- RAG
- Vector search
- Evaluation
- Open models
- TensorFlow
- scikit-learn
Languages 02
- TypeScript
- JavaScript
- Python
- Java
- C#
- SQL
Frontend 03
- React
- Next.js
- Angular
- Vue
- GraphQL
- Web components
Backend 04
- Node.js
- Django
- FastAPI
- .NET
- REST
- OpenAPI
- Stripe
- Serverless
Infrastructure 05
- AWS
- GCP
- Terraform
- Docker
- CI/CD
- PostgreSQL
- ClickHouse
- MongoDB
Education
B.A., University of RegensburgMedia Computer Science and Information Science, 2013 to 2017
Languages
German — first languageEnglish — level C1
Clients
Founders, product teams, enterpriseI work solo or inside your team.
04ApproachHow I work
The scope stays small.
The work stays visible.
The code stays yours.
- ScopeWe agree what is in and what is out before anything gets built. Short cycles keep the cost predictable and the exit cheap.
- EvidenceEvery AI claim gets measured before it gets a budget. If the numbers do not hold up, you hear it from me first.
- OwnershipYou own the code, the infrastructure and the documentation. Nothing runs on a service that only I can reach.
- ProgressYou see running software every week, not a status report about running software.
- LeadershipI can take the technical lead where you need one. I carried a startup's entire tech for four and a half years as co-owner and CTO — part-time works too.
Engagement/how
- Remote from Prague, CET
- German or English
- Day rate or fixed price
- From a two-day assessment
- To a project of some months
- Open for new projects
05ContactPrague · CET
