Tom Kisler

Building an end-to-end agentic AI voice assistant for deskless workers (warehouses, factory floors, logistics). Workers use push-to-talk to speak commands in plain language; an agentic AI workflow interprets intent and executes actions directly in WMS and ERP systems. Delivered a working prototype covering ASR backend (runs locally, no cloud API dependency), agentic orchestration layer, mobile and wearable apps, and an embedded hardware proof-of-concept. Prototype agent can look up SOPs and notify supervisors. Conducting customer discovery across industrial sectors.

Carbon13 is one of the world’s leading climate venture builders. Built and iterated AI-assisted and agentic product prototypes for a circular material selection concept. Led the preparation of investor materials: business model forecast, carbon case, roadmap, and data-room content. Cold outreach and customer discovery with ~50 potential customer conversations.

Co-led a €3M Product Carbon Footprint programme to automatically calculate product carbon footprints across the company, coordinating 10 part-time IT resources. Drove end-to-end introduction of CSRD-aligned sustainability reporting software: requirements engineering, provider selection, architecture feasibility, Betriebsrat approval, and budget monitoring. Led introduction of SiGREEN (Siemens PCF platform) and represented Wacker in the TfS (Together for Sustainability) working group. Among the first employees to receive a Microsoft Copilot licence, piloting enterprise GenAI use. Translated business needs into technical requirements and implementation roadmaps for 10+ enterprise transformation initiatives.

Built a co-founder matching platform from scratch. A Tinder-like web app where founders create profiles, like/pass on each other, take private notes, and chat live. Platform is live and actively used at recurring Munich meetup events. Implemented a state-of-the-art computer vision algorithm for an autonomous weeding robot (re-implementation from a research paper using GPU-enabled cloud compute). Additional ventures: AI-based ESG risk analysis for banks (concept & customer discovery), sharing platform (liquidated), sustainable fintech startup (pre-sales & platform concept, uncompensated exit).

Built a new internal business area developing a green service offering to support companies in their transformation towards more sustainable business models, processes, and software. Co-authored the Green Software Development Manifesto. Co-founded a Meetup for Green Software Development that merged with the Linux Foundation group on the same topic, making it the second-largest group worldwide.

Acquired and technically led the early phase of a publicly funded AI research project (RobIn4LeMi). Led a confidential AI/ML pre-development project at BMW, aligning customer needs with feasible technical solutions and a 5-year roadmap. Assessed performance-test quality on BMW’s high-performance infrastructure for autonomous driving data processing. Led team development of an internal web application. Voted Top 5 / Most Innovative at MAN Truck & Bus Hackathon (jury of MAN, Microsoft, and digitalisation experts).

Developed full-stack research software (WebMAUS / BAS Web Services) within the European CLARIN infrastructure initiative. The platform receives 5M+ API calls per year and is widely used in the international speech-research community. Presented at approximately 20 international conferences and workshops. Published 14 peer-reviewed papers with 1,200+ citations. Taught 7 undergraduate and graduate courses.

Doctoral research on large-scale speech data analysis using machine learning (Random Forests, SVR, SVM). Localised speakers’ regional origins from acoustic features; validated and error-checked automatic pre-processing methods for speech recordings; related findings to phonetic and dialectological knowledge. Dissertation: Methods for Large-Scale Data Analyses of Regional Language Variation Based on Speech Acoustics (magna cum laude).

Organising two recurring Munich meetup formats connecting potential co-founders (general and impact-focused). Built a dedicated matching platform. A Tinder-like web app where founders create profiles, like/pass on each other, take private notes, and communicate via live chat. The platform is live and actively used by meetup attendees.

Active since foundation (March 2019, initial signatory). One year each in the S4F Munich coordination team and the national working group ‘Industry and Companies’ (lead of the ‘Networking’ sub-group).