Robert Söldner

Building with AI, and checking that it works.

Software engineer with a decade of experience across backend systems, developer tooling and applied AI. Founder of wuselfabrik, and completing a PhD on the mechanised verification of graph programs.

Writing

Experience

Founder at wuselfabrik GmbH

2024–now
  • Founded a consulting vehicle for software engineering and applied-AI work, focused on LLM- and voice-based automation: document understanding, conversational agents, and voice generation on open-source and hosted foundation models.
  • Delivered client engagements spanning backend services, data pipelines, and ML/LLM integration into existing product workflows.

Software Engineer at Cantor8 Technologies S.A.

2025–2026
  • Developed Canton smart contracts in Daml for institutional tokenised-asset workflows on the Canton Network.
  • Extended and maintained the Python wallet backend supporting asset custody and transaction flows.

Software Engineer at Kadena LLC

2022–2025
  • Co-authored a complete rewrite of the Pact smart-contract compiler, achieving 8x on-chain storage efficiency and 2–3x execution speedup, cutting the cost of block processing across the network.
  • Designed and implemented a Language Server Protocol backend from scratch (jump-to-definition, autocomplete, execution-trace highlighting), shortening smart-contract development cycles across the ecosystem.
  • Prototyped a zero-knowledge bridge between Ethereum and Kadena using Succinct’s SP1 framework, demonstrating cross-chain interoperability and informing the product roadmap.
  • Expanded automated correctness checking for user-written smart contracts, significantly reducing the security review burden for contracts deployed to production.

Senior Scientist at Sartorius Stedim Biotech GmbH

2018–2022
  • Established a research line at the interface of AI, automation, and bioprocessing, resulting in 7 peer-reviewed publications and 10 patent filings (3 granted, 7 pending) covering bioprocess monitoring, sterility testing, and integrated continuous manufacturing.
  • Developed a machine-vision system for real-time foam detection in bioreactors and transitioned the prototype to a product team, enabling automated process monitoring in production.
  • Contributed to spectroscopy-based inline monitoring of CHO perfusion cultivations, including deployment of O-PLS models for viable-cell-concentration estimation.
  • Built automation tools and proofs-of-concept for visual programming of bioprocess data and smart-laboratory device interaction on state-of-the-art bioprocessing equipment.
  • Mentored IT apprentices and cross-functional teams on software engineering practices.

Software Engineer at 2-FIFTY-5 Amaizing Solutions GmbH

2015–2017
  • Led a team of four engineers to architect and deliver a cloud-based analytics platform serving customer projects across resource planning, healthcare, and sales forecasting.
  • Evolved the analytics engine from a single-core prototype into a distributed multi-core architecture with on-demand cluster provisioning, cutting query times from hours to minutes.
  • Managed a fully-distributed remote engineering team; established code review, workflows, and CI/CD pipelines.

Education

PhD, Computer Science at the University of York 2020–expected 2026/27
Thesis (working title): A Verification Assistant for Graph Programs. Supervisor: Prof. Detlef Plump. Mechanisation of the double-pushout approach to graph transformation in Isabelle/HOL and HOL4.
B.Sc., Computer Science at IUBH University of Applied Sciences 2019–2020
Thesis: Implementation of a data handling toolbox in compliance with “good practice” quality guidelines and regulations (GxP).

Awards & boards

Research