Benedikt Engel
About me

Hi, I'm Benedikt Engel. I'm currently pursuing my Master's degree in Computer Science at the Karlsruhe Institute of Technology Top-tier technical university in Germany: kit.edu, with a strong focus on Machine Learning, Data Science, and Robotics. I thrive at the intersection of applied AI and software engineering. Working alongside my studies as a working studentA working student is a student who, in addition to their full-time studies at a university, is employed by a company on a more-than-marginal basis., my journey has taken me from developing robotics software at our Robotics Institute High Performance Humanoid Technologies Lab (H²T), to building automated data pipelines as a Research Analyst at Horváth & PartnersTop-tier German Management Consulting: Horváth, to my current role engineering scalable, AI-driven applications as a Full-Stack Developer at LexaruSuccessful Legal-Tech Start-Up: Lexaru. Whether I'm researching contextual multi-armed bandits, launching entrepreneurial side-projects like TableXAI, or designing fluid web architectures, I love building things. Let's make it happen!

I am always open to exchanging ideas. If you would like to discuss something, please don't hesitate to reach out!

Languages

  • Python Used in: TableXAI, Sourcerer, RAG Chat, Sentiment Analysis, Fluxio, Tiny YoloV2
  • Typescript Used in: Full Stack Developer @ Lexaru
  • C++ Used in: B.Sc. Computer Science @ KIT
  • Java Used in: B.Sc. Computer Science @ KIT
  • Haskell Used in: B.Sc. Computer Science @ KIT

Frontend

  • Svelte Used in: Full Stack Developer @ Lexaru
  • Vue.js Used in: TableXAI MVP Web Service

Backend & DB

  • PostgreSQL Used in: TableXAI Account Management
  • MongoDB Used in: Various Web Projects
  • Drizzle ORM Used in: Full Stack Developer @ Lexaru
  • Chroma DB Used in: Ollama Mini RAG Chat Vector Storage

ML & Data

  • PyTorch Used in: Tiny YoloV2 on Jetson Nano, Budgeted Contextual MAB
  • TensorFlow Used in: Deep Learning Coursework
  • PaddleOCR Used in: TableXAI Local Extraction Pipeline
  • RAG Used in: Ollama Mini Chat System

Systems

  • Linux Used in: H²T Lab Robotics, Jetson Nano Deployment
  • MacOS Used as: Primary Development Environment
  • Windows Used for: Everyday Tasks
  • Docker Used in: Containerized Deployments & Local Dev

Experience & Projects

April 2026 - Present

Full Stack Developer @ Lexaru

Developing AI applications to automate workflows for authorities, engineering scalable architectures using Svelte, TypeScript, and Drizzle ORM.

June 2026 - June 2026

ML Hackathon: Sensor Data Classification @ AIMAT (KIT)

Participated in a competitive university hackathon focused on classifying complex, heavily distorted sensor data using ResNet18 (preTrained then finetuned). Engineered features and trained a robust model under strict time constraints without using guided templates.

  • Achieved >88% overall accuracy on the hidden test set, clearly outperforming the baseline.
  • Techniques: Ensemlbe, TTA, Image Preprocessing, Data Augmentation
  • Secured bonus placement based on model performance metrics.

April 2026 - August 2026

Seminar Paper on Regret Proofs of Multi Armed Bandit Algorithms @ IPD (KIT)

Rederived wort-case regret bounds for the Upper Confidence Bound and Linear Upper Confidence Bound algorithms.

  • Regret Proof for the (Linear) Upper Confidence Bound Algorithm, showing sub-linear Regret.
  • Empirical evaluation on the Movielens 1M dataset using rejection sampling.
  • Graded 1.0.

Oct 2025 - March 2026

Hardware efficient Offline Person Detection w/ Tiny YoloV2 on Jetson Nano @ CES (KIT)

Trained, Fine-tuned and deployed Tiny YoloV2 for offline Person Detection on a Jetson Nano.

  • Pretrained on ImageNet.
  • Finetuned using ablation method with progressive unfreezing of layers, to preserve low level feature.
  • Optimization using int8 and float16 Quantization with per Tensor Quantization.
  • Graded 1.0.

Oct 2025 - Jan 2026

Founded TableXAI - Table extraction Web application

Extract tables from images and PDFs using PaddleOCR and Mistral OCR.

  • Complete user management and account management.
  • Local extraction pipeline able to run on Server CPU.
  • Mistral OCR for more complex tasks.
  • https://tablexai.de

October 2025 - Present

M.Sc. Computer Science @ KIT

Master's degree in Computer Science with focus on ML, Data Science and statistics.

Oct 2024 - April 2025

Bachelor Thesis @ IPD (KIT)

Developed novel contextual multi-armed bandit algorithms under fixed budget constraints.

  • Integrated Neural Networks, Gaussian Processes, and Ridge Regression into the decision-making loop.
  • Graded 1.0.

April 2025 - April 2026

Research Analyst @ Horváth

Data-driven analysis with a focus on the automotive sector.

  • Developed automated web scraping pipelines.
  • Built interactive dashboards for sentiment analysis.
  • Active market analysis and research
  • First 7 month internship followed by 5 months as a working student.

April 2024 - August 2024

Seminar Paper on Software Sustainability @ SDQ (KIT)

Review-Paper about impacts on software on the environment focusing on making software developmenent more environmentally friendly.

  • Overview of negative impacts of software on the environment.
  • Summary of Frameworks for efficient software devlopment, efficient cloud design, Cyber-foraging, Sustainable Reference Architectures and Green IT.
  • Graded 1.7.

March 2023 - Oct 2024

Assistant Researcher @ H²T Lab (KIT)

Co-founded and developed Fluxio, a web-based statechart editor designed to simplify robot programming for non-experts.

October 2021 - September 2025

B.Sc. Computer Science @ KIT

Bachelor's degree in Computer Science with focus on theory, ML and math.