Founder-CTO · Enterprise AI · Technology leadership
How I got here, and how I work.
Co-founder & CTO of Jaden Data until June 2024 · fractional CTO and consulting partner since. We bootstrapped the company past €1M in annual revenue with no external capital; I grew engineering from 0 to 10, took it through ISO 27001 and SOC 2 Type 2 in about three months, and ran a platform that served 50+ organisations at 99.9% uptime.
Open to CTO and Head of AI roles · remote-first, on-site when it matters
Background
The route I took
I started out where mistakes are expensive. At SAS Institute I worked on a real-time behavioural analysis project for ING, designing pipelines that handled thousands of events per second on Kafka, Flink and Camel — one of the first SAS Event Stream Processing implementations in the European Union — and I wrote a data-lineage plugin for GDPR reporting that analysts across ING Group ended up using. At Deutsche Telekom's T-Labs I designed and deployed a distributed Hyperledger Fabric network for package tracking, with the chaincode, the secured REST API and the identity service around it. At idealo I was a key contributor to the six-country login rollout that created 1.4M accounts in three months, 700% above forecast, and I carried on-call duty for services that were not allowed to fail.
In 2021 I left to co-found Jaden Data, and was its CTO until June 2024. We bootstrapped it — no external capital — past €1M in annual revenue, and I grew engineering from 0 to 10 across frontend, backend and AI. I owned the technology strategy and the product roadmap, the architecture of our platform and of the client systems built on it, the technical budget, security and compliance, technical pre-sales and project kick-offs, and the reporting that went to clients and shareholders. The platform ran for 50+ organisations at 99.9% uptime, and we passed ISO 27001 and SOC 2 Type 2 in about three months without pausing delivery. Not all of it was AI: the same team delivered regulated distributed systems, including a cross-ledger CBDC bridge with Norges Bank and research platforms funded under EU Horizon 2020.
Since June 2024 I have been Jaden Data's fractional CTO and consulting partner, and I take engagements directly: interim technology leadership, architecture and risk reviews, production AI for document-heavy workflows and customer-facing assistants, and getting engineering teams to work with AI tools well and securely. I still write code — RP-Matcher, the machine-learning matcher that cut offer preparation time by 60% for RP Group, and Taufolio, which I am building now. What I want next is the permanent version of the job I already did: a CTO or Head of AI role where the technology, the team and the outcome sit with the same person.
How I operate
Four things that do not change
-
Start with the workflow, not the model
Before anything is designed I sit with the people doing the job and map the process, the baseline, the bottleneck, the risk and the economics. At a German development bank that produced one specification template and one evaluation template used across a portfolio of proofs of concept, so results could be compared instead of argued about.
-
Ship the smallest loop that really runs
One end-to-end path through real data, real systems and real governance beats a broad prototype. At APG the first assistant was assembled in a day from platform components; the twelve sprints that followed went into integration, user validation and the on-premises deployment.
-
Treat evaluation, security and observability as product capabilities
They belong in the build, not in a phase at the end: deterministic and model-graded test sets, retrieval metrics, error taxonomies, human-review thresholds, role-based access control and audit trails. ISO 27001 and SOC 2 Type 2 were run as an engineering programme alongside delivery, not as a document exercise.
-
Leave a stronger team and a reusable system
Every deployment should make the next one cheaper: shared components, evaluation harnesses, delivery standards and written decisions. Code review, CI/CD quality gates and testing requirements were in place before the team grew, so quality never depended on one person.
Leadership scope
What I have owned
Not an aspiration list. Each of these was mine to decide, budget, staff and answer for.
-
Strategy and roadmaps
Defined and ran the technology strategy and product roadmap at Jaden Data, aligned with revenue targets and the delivery calendar, and decided what the platform would and would not do.
-
Technical pre-sales and customer architecture
Discovery, executive workshops, demos, architecture and security reviews, scoping and business cases. Four procurement workflows at a global automotive group went from scoped use case to production this way.
-
Team building and mentoring
Hired and led engineering from 0 to 10 across frontend, backend and AI workstreams, with structured code review, pair programming and internal knowledge sharing.
-
Engineering standards
Code-review rules, CI/CD quality gates and testing requirements set before the team grew, plus the redesign of an entangled frontend monolith into micro-frontends so teams could deploy independently.
-
Technical budgets
Owned the technical budget: cloud spend, third-party model and API costs, and resourcing across parallel client deliveries in a company with no external funding.
-
Risk and compliance
ISO 27001 and SOC 2 Type 2 in about three months, multi-tenant isolation and role-based access control, GDPR, on-premises and data-sovereign deployments, security audits and failover planning.
-
Executive communication
Reporting to clients, shareholders and partners. At a global automotive group I raised the stakeholder cadence from bi-weekly to weekly and split it into parallel sub-groups, so dependencies stopped blocking delivery.
-
Production delivery
Accountable for systems in production, not pilots: 50+ organisations on one platform at 99.9% uptime, an on-premises MVP for APG, four tendering and contracting workflows live at a global automotive group.
Technical depth
What I can still do myself
-
AI systems
Multi-step LLM processes on an event-driven platform: parallel extraction into JSON, reconciliation, then generation with source references preserved; retrieval with reranking and page-level citations; structured-output prompting with strict language control; MCP-based tools exposed to agent clients.
-
Evaluation
Testbed, the framework I designed for AI processes and retrieval pipelines: deterministic test cases, quantitative and qualitative scoring, error taxonomies, human-in-the-loop validation, and launch criteria that weigh quality against latency and cost.
-
Cloud and distributed systems
Event-driven microservices on AWS (Lambda, ECS/Fargate, SNS/SQS, CloudWatch) behind a multi-tenant platform with role-based access control, over 1,000 concurrent connections and thousands of requests per second at 99.9% uptime, with a provider-independent model layer across OpenAI, Azure, AWS Bedrock, Google Gemini and Mistral.
-
Data, streaming and machine learning
Kafka, Flink and Camel for real-time processing; PostgreSQL, pgvector, Pinecone and Milvus for retrieval; and classical machine learning where it wins — the tuned RP-Matcher ensemble reaches 78.4% top-1 accuracy on the current corpus, on 25+ engineered features.
-
Security and compliance engineering
ISO 27001 and SOC 2 Type 2, multi-tenant isolation and role-based access control, GDPR and data lineage, on-premises deployment under financial regulation, audit trails and security reviews.
Timeline
Roles, and what came out of them
One business outcome and one technical contribution per role.
-
Jun 2024 – present
Fractional CTO & consulting partner Jaden Data and direct client engagements
RP-Matcher cut offer preparation time by 60% for RP Group and removed the dependency on one expert for product selection.
Built and benchmarked five model families with hyperparameter search; the tuned ensemble reaches 78.4% top-1 accuracy on the current corpus, on 25+ engineered features.
-
Mar 2021 – Jun 2024
Co-founder & CTO Jaden Data
A bootstrapped company at €1M+ annual revenue with 50+ organisations on the platform, ISO 27001 and SOC 2 Type 2 in about three months, and engineering grown from 0 to 10.
Architected the multi-tenant, provider-independent AI platform on event-driven AWS, and led the redesign of an entangled frontend monolith into micro-frontends.
-
Apr 2020 – Jun 2021
Software Engineer idealo
Key contributor to the six-country login rollout: 1.4M accounts in three months, 700% above forecast.
Real-time services on Kafka, WebSockets and NoSQL, production microservices and REST APIs in Java and Kotlin with Spring Boot, and on-call duty for 24/7 availability.
Employer reference (PDF) -
May 2019 – Apr 2020
Blockchain Engineer Deutsche Telekom Innovation Laboratories (T-Labs)
A package-tracking system with an immutable shipment record that partner mobile apps and ERP systems could both work against.
Designed the architecture, deployed a distributed Hyperledger Fabric network with chaincode, and built the secured REST API, the proxy service and the authentication service around it.
-
Sep 2016 – Sep 2018
Associate Technical Consultant SAS Institute · banking clients including ING
A GDPR data-lineage plugin adopted by analysts across ING Group internationally, and one of the first SAS Event Stream Processing implementations in the European Union.
Designed and implemented real-time decisioning pipelines handling thousands of events per second on Kafka, Flink and Camel.
On paper
Education, awards and publications
Education
-
M.Sc. Computer Science
Distributed systems, security and machine learning; thesis on blockchain-based monetisation of digital content.
-
M.Sc. Computer Science
Double-degree programme; honours diploma.
-
B.Sc. Electronics and Computer Engineering
Electronics, programming and computer architecture, with an engineering-practice focus.
Awards
-
Second place, Berlin Legal Tech Hackathon
Led the team that built a prototype for detecting and moderating online hate speech, combining language processing with real-time content analysis.
-
University President's undergraduate scholarship
Awarded four years in a row for academic performance during my undergraduate studies.
Publications
-
KnowledgeX: Trusted Inter-organizational Data Processing
-
DeCoCo: Blockchain-based Decentralized Compensation of Digital Content Purchases
-
Saluki: Finding Taint-style Vulnerabilities with Static Property Checking
Off the clock
Outside work
I spend my free time in the air and in the water: certified skydiver (PJ(B) licence), Advanced Open Water diver, International Windsurfing Instructor, and a snowboarder whenever there is snow.
I speak Polish natively, English fluently and German conversationally, and I work remote-first, on-site when it matters.
Two ways to work with me
Hiring a technology leader, or building something that has to run in production? Pick the door that fits.
Fractional CTO & consulting partner · open to the right permanent CTO or Head of AI role