Dennis Dickmann-Founder
Check rate
Experience
Founder
Latence
- Founded Latence to commercialise runtime safety patterns from HALO as a deployable product.
- Built end-to-end as single technical founder with open-source stack on NVIDIA ecosystem.
- Developed TRACE: real-time safety layer for knowledge agents and RAG pipelines with groundedness scoring, prompt-attack detection, GDPR redaction, context compression, audit-ready traces.
- Developed vLLM Factory: production inference framework on vLLM with custom Triton kernels and 12 parity-validated plugin models, achieving up to 11.7× throughput vs vanilla PyTorch.
- Developed ColSearch: single-node multi-vector late-interaction retrieval engine with Rust SIMD and fused CUDA, achieving 3.12× FastPlaid geomean QPS on BEIR-8 and a 1.58-bit quantized lane 6.4× smaller than FP16.
- Developed llm-opt: LLM compression research framework with hierarchical importance, structured pruning, tabu search, knowledge distillation.
Co-Founder & CTO
seedbox.ai
- Co-founded the company in 2020 and closed €500k pre-seed from a Stuttgart family office.
- Built seedbox into a recognised German AI lab serving 25+ enterprise customers across DACH.
- Built and led the engineering organisation.
- Act as the senior technical counterpart on every customer and partner engagement: owning technical discovery, executive workshops, solution architecture, evaluation design, proof-of-value, deployment-pattern selection, and the path from first conversation to production handover.
- Delivered production LLM systems for Landesbank Baden-Württemberg, WAGO, fischerwerke, Bertrandt, Reguvis Fachmedien. Selected systems: multi-vector medical RAG, secure on-prem private chat for a financial institution, B2B sales-research agent for complex manufacturing, Velaro real-estate appraisal co-pilot, German-language continuous pre-training of state-of-the-art open-source LLMs.
- Owned the pre-sales process end-to-end: customer technical discovery, complex buying-cycle navigation across IT, legal, procurement, executive sponsors, evaluation framework design, deployment-pattern selection (cloud, VPC, on-prem, air-gapped), and integration with existing enterprise systems (IAM, data boundaries, observability).
- Flagship project HALO: principal investigator on a €200k EuroHPC research grant (FFplus, 2024–2026), delivering ~85% operational cost reduction and sentence-level auditability.
- Established strategic technology partnerships with HPE, AMD, and HLRS.
- Hosted Boost with AI podcast, moderating technical-strategic conversations on enterprise AI, the EU AI Act, prompt-injection safeguards, and LLM safety.
Head of Product
Lab1886 — Mercedes-Benz corporate incubator
- Led the Stuttgart product organisation of Daimler's corporate startup factory Lab1886.
- Built and scaled product practice across venture portfolio, taking digital products from ideation through pilots to operational deployment.
- Hired, developed and led product managers across multiple ventures in mobility services, B2B offerings, customer-data initiatives and adjacent digital businesses.
- Navigated multi-stakeholder environments across business units, customer pilots, IT, legal, executive sponsors and external partners; translated product capabilities for executive and engineering audiences.
Lead Product Manager
Lab1886 — Mercedes-Benz corporate incubator
- Owned product management for digital business-model initiatives from problem definition and customer-discovery through MVP delivery and pilot operations.
- Collaborated cross-functionally with Daimler business units, legal, technical implementation teams, customer pilots and external partners.
Manager Advanced Analytics & Big Data
Daimler Truck AG
- Owned advanced-analytics and big-data initiatives inside Daimler Truck AG.
- Gained hands-on exposure to large-scale industrial data systems, enterprise IT adoption, and stakeholder maps for internal data programmes.
Industry Experience
See where this freelancer has spent most of their professional time.
Experienced in Information Technology and Automotive.
Business Area Experience
See which departments and functions this freelancer has contributed to most.
Experienced in Product Development, Information Technology, Research and Development, and Business Intelligence.
Summary
Customer- and partner-facing technical leader with 10+ years architecting AI and HPC solutions for enterprises and ecosystem partners across DACH. As Co-Founder & CTO at seedbox.ai I serve as the senior technical counterpart to 25+ enterprise customers (SAP, LBBW, WAGO, fischerwerke, Bertrandt, plus financial services and publishing) and to OEM, silicon and HPC partners (HPE, AMD, HLRS). Used to own engagements from technical discovery through architecture, evaluation and production deployment. Work spans LLM training and distillation on EuroHPC infrastructure, production GPU inference (vLLM, CUDA, Triton), and deployment from managed cloud to fully air-gapped. Equally at home in boardrooms, engineering deep dives and partner workshops.
Skills
Communication & Customer-Facing: Executive Briefings · Engineering Deep Dives · Partner Enablement Workshops · On-Camera Customer Stories · Conference Stages · Podcast Hosting · Business-Technical Translation · Channel & Oem Partnership Development.
Pre-Sales & Solutioning: Llm Discovery · Solution Architecture · Gpu Evaluation Framework Design · Proof-Of-Value · Technical Training Delivery · Ecosystem & Community Building.
Hpc & Accelerated Computing: Distributed Training On Eurohpc-Class Infrastructure (Hlrs) · Multi-Gpu Clusters · Cuda · Triton · Nccl · Profiling.
Llm & Ml Systems: Continuous Pre-Training · Rl · Knowledge Distillation · Pruning · Quantization.
Engineering & Deployment: Python · Pytorch · Vllm · Fastapi · Rust · Cloud (Aws, Gcp, Azure) · Vpc · On-Prem · Iam · Observability · Gdpr & Eu Ai Act.
Production Llm Inference On Gpu Infrastructure — Vllm-Based Encoder Serving Framework With Custom Triton Kernels, Up To 11.7× Throughput Over Vanilla Pytorch On Identical Systems Hardware: Vllm Factory.
Multi-Vector Retrieval At Production Scale — Rust Simd With Triton And Fused Cuda Kernels Optimized Search Engine With Novel 1.58-Bit Riemannian Quantization Lane: Colsearch.
Llm Training And Distillation On Eurohpc Infrastructure — Continuous Pre-Training, Grpo Reinforcement Learning, Knowledge Distillation Across 100,000+ Gpu Hours On Hlrs Clusters. Peer-Reviewed At Cug '25 (Acm).
Production Rag Over Proprietary Corpora — Medical, Financial, Legal — With Sentence-Level Groundedness Scoring And Hallucination-Audit Frameworks Built In.
Multi-Tier Enterprise Deployment — Managed Cloud, Customer Vpc, On-Prem And Fully Air-Gapped — Same Api And Contract At Every Boundary.
Enabling Adoption Through Workshops, Training & Ecosystem Building — German-Language Technical Workshops With Executive And Engineering Audiences; Partner-Enablement Sessions; Host Of A Public Podcast Moderating Discussions With Senior Enterprise Leaders And Ai Researchers.
Languages
Education
University of Bremen
Diploma in Industrial Engineering & Management · Industrial Engineering & Management · Bremen, Germany
Statistics
Experience
Global Experience
Expertise
Qualifications
Profile
Frequently asked questions
Have questions? Find more information here.
Average rates for similar positions
Rates are based on recent contracts and do not include FRATCH margin.
Similar Freelancers
Discover other experts with similar qualifications and experience
Jürgen Fey
Ralph Navasardyan
Ronald Foerster
Ulf Schiebener
Matthias Isler
Martin Musiol
Marcel Meyer
Laurin Hagemann
Yannick Schuchmann
Paul Webster
Experts recently working on similar projects
Freelancers with hands-on experience in comparable project as a Founder
Ralph Navasardyan
Jürgen Fey
Mahabub Akram
Ursula Maria mayer
Martin Musiol
Stephan Fröde
Alexander Schulze
Yannick Schuchmann
Matthias Isler
Nima Nooshi
Nearby freelancers
Professionals working in or nearby Stuttgart, Germany
Cornelia Ilg
Carlos Medina
Robin Sadlo
George Ojog-Schulze
Thomas Gaiser
Florian Becker
Lars Otto
Elvir Büyükbayrak
Hasan Aydin
Will Orrantia
Most recent projects
FRATCH works with many companies and recruitment agencies. Here you will find our recently posted projects and opportunities.
