Profile

Lukas Döllerer

M.Sc. Student @TUM • Software Engineer

About Me

I'm a master's student @TUM studying Informatics and a software engineer with over 5 years of professional experience in web- and backend development. I specialize in the building of high performance compilers and runtime systems for programming languages. Currently focused on high speed execution of WebAssembly and an alternative Django webserver runtime.

Experience

Software Developer

CedarDB January 2026 - Present

Student Researcher

TUM Chair for Database Systems April 2025 - December 2025

  • Worked on LingoDB, a compiler-based analytical database system built on MLIR.
  • Implemented a novel backend using tpde, which translates MLIR directly to x86_64 assembler, skipping LLVM compilation for significantly lower execution latency.

Software Developer in R&D

Rohde & Schwarz GmbH & Co. KG February 2023 - March 2025

  • Developed C++ libraries to enable high-speed communication between system components of signal analyzers
  • Implemented diverse communication protocols, including TCP, HTTP, MQTT, ZMQ, and shared memory

Software Developer for Internal Services

CHECK24 Services GmbH April 2021 - Januar 2023

Developed and deployed multiple Python Django web services, integrated with JavaScript frontends for HR

Software Developer for Internal Services

IfTA Systems GmbH June 2020 - October 2021

Engineered a web-based time tracking system for HR, streamlining employee productivity management

Projects

Physics Exams Platform (MPIC TUM) logo

Physics Exams Platform (MPIC TUM)2021 - Present

  • Developed a web platform for sharing and searching physics exams and solutions for TUM students.
  • Features full-text search, PDF previews, and a modern, mobile-friendly UI.
  • Used by hundreds of students each semester to prepare for exams.
Next.jsTypeScriptTailwind CSSPostgreSQLDocker
LingoDB - MLIR Database System logo

LingoDB - MLIR Database SystemOngoing Contribution

  • Contributed to the development of LingoDB, a novel database system leveraging compiler technologies (MLIR) for high-performance analytical data processing.
  • Involved in aspects of query optimization and system architecture.
  • LingoDB aims to bridge the gap between database systems and compiler frameworks for flexible and efficient data anlysis.
C++MLIRDatabase SystemsQuery Compilation
Efficient Point Cloud Streaming for EDGAR TUM Self-Driving Car  preview

Efficient Point Cloud Streaming for EDGAR TUM Self-Driving Car January 2025 - Mai 2025

  • Developed efficient pointcloud compression techniques for autonomous driving vehicles
  • Created solutions for HD map updates with optimized data streaming
Google DracoCompressionAutonomous DrivingHD Maps

WebAssembly PGOOngoing Contribution

  • Bringing high performance execution to WebAssembly through the magic of profile-guided optimization (PGO) 🚀
WebAssemblyLLVM

WebAssembly Standalone RuntimeApril 2024 - December 2024

  • Created a fully spec-compliant runtime featuring AOT & JIT compilation using LLVM
  • Conducted research on efficient bounds checking, leading to a paper published at the VMIL 2024 conference
RustLLVMWebAssembly

A Universal Serialization Framework for C++February 2023 - December 2023

  • Invented an automated (de-)serialization system using template metaprogramming for generic C++ objects
  • Designed an extendable framework with support for multiple serialization formats
  • Implemented support for JSON, MessagePack, Protobuf, Thrift, and FlatBuffers protocols
C++Template MetaprogrammingJSONMessagePackProtobufThriftFlatBuffers

Static Binary Translation from RISC-V to x86_64April 2021 - October 2021

  • Developed a static binary translator to convert RISC-V binaries to x86_64
  • Achieved faster performance than dynamic binary translators like QEMU
C++RISC-Vx86_64Binary TranslationQEMU

Papers & Theses