Florian
Zeba.

Consulting and building, at the same time — not in sequence.

A CV flattens a career into a list, which hides the part that matters. The consultancy roles, the engineering work and my own practice were not phases. They ran together — FZEBA.com has been open alongside every employed role since February 2025.

Where the work was done.

The chart above names them; this says what they are. Each row states the actual relationship — payroll, advisory, or a company of my own.

• ORBIS Austria

  2026 — present

  Employment

  Enterprise data and AI architecture

  A European business consultancy and systems integrator working across SAP, Microsoft and manufacturing IT. As Principal Enterprise Architect for Data & AI I own the platform and AI strategy: what gets built, on which stack, and how governance, security and scale are designed in rather than retrofitted.

  orbis.de ↗

• CANCOM Austria

  2025 — 2026

  Employment

  Data and AI platforms for enterprise clients

  One of Austria's larger IT service providers. Solutions architect in the Digital Makers team, advising enterprise architects, solution designers and C-level on data and AI architecture, and delivering data-intensive business cases end to end — from the first requirements workshop to a running platform.

  cancom.at ↗

• FZEBA.com

  2025 — present

  Founded & led

  My own name, my own engagements

  Data-driven software development taken on directly rather than through an employer, and the vehicle for work that does not fit inside a consultancy engagement. It has run alongside every employed role since 2025.

  fzeba.com ↗

• SOFTWERK

  2025

  Employment

  Data engineering for finance, manufacturing and agriculture

  A software engineering firm building custom systems for Austrian mid-market clients. Led data engineering projects from requirements to delivery, and automated pipelines in Azure Data Factory with PySpark that replaced manual data handling in finance, manufacturing and agriculture.

  softwerk.at ↗

• PwC Österreich

  2022 — 2025

  Employment

  Management and BI consulting, then tax technology

  The Austrian arm of the professional services firm. Two years and eight months across two roles: first as Senior Associate building automated data pipelines and BI dashboards for controlling and logistics, then in Tax Technology, where the work was automating tax, accounting and controlling processes with RPA and making ERP data — BMD, Business Central, Dynamics — actually usable for analytics.

  pwc.at ↗

Governance you can prove, not guardrails you hope hold.

Cortex: A Fixed-Point Theory of Governed Coding Agents — co-authored with Marius-Constantin Dinu at Alpha Omega Labs. It treats an agent's validate–repair loop as a monotone operator on a lattice of satisfied requirements, so “done” becomes a least fixed point you can prove rather than a heuristic stopping rule.

• 2026-08-10 · 28 min The Graph Is Not the Point: Building a Personal Knowledge System That Actually Teaches You Something Why building a concept map beats studying one, why the untyped wikilink throws away the information you needed, and how to layer embeddings, LLM triple extraction, and structural gap detection on top without letting the machine author your graph.
• 2026-06-05 · 20 min Building a Write-Back Interface on Microsoft Fabric Apps with Rayfin A full technical walkthrough of building a governed write-back interface on Microsoft Fabric Apps with the Rayfin SDK — static hosting, Entra ID SSO, and a server-side functions proxy to an existing REST API, with no separate web app to maintain.
• 2026-05-29 · 9 min Stop the Garbage Before It Lands: A Deep Look at Databricks Labs DQX A practical guide to DQX, the PySpark-native data quality framework from Databricks Labs that quarantines bad data before it reaches your gold tables.
• 2026-05-23 · 8 min Multi-Agent Pipelines for Academic Figures: The PaperBanana Open-Source Implementation A technical deep dive into PaperBanana, an open-source agentic pipeline that generates publication-quality methodology diagrams and statistical plots from text.
• 2026-05-18 · 9 min Stop Tinkering With Strings: A Practical Tour of DSPy and Reflective Prompt Evolution Why Stanford's DSPy framework treats prompts as compiled artifacts, and how GEPA — the ICLR 2026 Oral — outperforms reinforcement learning with 35x fewer rollouts.
• 2026-05-13 · 8 min Beyond Tidy Data: How Microsoft Data Formulator Lets AI Reshape Your Way to a Chart Exploring Microsoft Research's Data Formulator, the concept-binding paradigm, and how its AI agents remove the tidy-data tax from visualization authoring.