Portrait of Lucian Dorin Crainic wearing a dark suit and tie
Rome, Italy

About

Lucian Dorin Crainic

Software Development Engineer in Test

Computer Science MSc Student

Software Development Engineer in Test specializing in embedded systems testing, with a Computer Science background and a strong interest in reliability, performance, engineering efficiency, continuous learning, and emerging technologies.

I design and maintain automated test systems for embedded software, turn requirements into traceable verification work, and collaborate across software, firmware, quality, and system engineering.

Current work

Building confidence in embedded systems.

I design and maintain automated test systems for embedded software, turn requirements into traceable verification work, and collaborate across software, firmware, quality, and system engineering.

ELT Group

Software Development Engineer in Test

Rome, Italy · – Present

MSc and research focus

Measuring how architecture shapes inference.

My current academic work investigates how CPU architecture and memory bandwidth shape LLM inference performance, with a focus on CPUs equipped with High-Bandwidth Memory.

Current research

LLM Inference on CPUs with High-Bandwidth Memory

High Performance Computing

M.Sc. in Computer Science · Sapienza University of Rome

Experience

Verification as an engineering system.

Requirement-driven testing, reusable automation, and collaboration across software and hardware disciplines.

ELT Group

Software Development Engineer in Test

Rome, Italy

Present

As a Software Development Engineer in Test at ELT Group, a leader in advanced Electronic Warfare systems, I help ensure the reliability, performance, and compliance of mission-critical embedded software powering next-generation RESM sensor products.

Responsibilities

  • Define, document, and implement Test Plans and Test Cases in compliance with MIL-STD-498 and DO-178C standards, ensuring traceability and verification coverage for safety-critical systems.
  • Design, develop, and maintain automated test frameworks for embedded software in Electronic Warfare sensor systems using Python, C++, and Robot Framework.
  • Validate RESM system functionality, including ESM and ELINT capabilities, through scenario-based and requirement-driven testing.
  • Verify and assess Built-In Test functionality, including CBIT, IBIT, and PBIT, across operational states and diagnostic conditions.
  • Develop and manage CI/CD pipelines using GitLab, Docker, and Kubernetes to streamline test execution, enable parallel validation, and improve software delivery efficiency.
  • Collaborate with Software, Firmware, QA, and System Engineering teams to validate software-hardware integration and compliance with system-level requirements.
  • Design and maintain reusable and scalable test suites for RESM products, supporting standardized testing approaches across multiple programs and configurations.

Technical terminology

RESM
Radar Electronic Support Measures
ESM
Electronic Support Measures
ELINT
Electronic Intelligence
BIT
Built-In Test
CBIT
Continuous Built-In Test
IBIT
Initiated Built-In Test
PBIT
Power-On Built-In Test

Education

Computer science, systems, and research.

A continuous path from machine learning foundations to systems, networking, and high-performance computing.

  1. 01

    Master of Science

    M.Sc. in Computer Science

    Sapienza University of RomeCurriculum: Systems and Networking
    High Performance Computing

    Thesis: LLM Inference on CPUs with High-Bandwidth Memory

    Present

  2. 02

    Bachelor of Science

    B.Sc. in Computer Science

    Sapienza University of Rome
    Machine Learning

    Thesis area: Machine Learning

    View thesis repository

Technical skills

Tools grouped by how I use them.

CV-derived skills are kept distinct from tools currently used in MSc and thesis work. No proficiency rankings are implied.

CV background01

Programming languages

  • Python
  • C++
  • Bash
CV background02

Frameworks and libraries

  • Robot Framework
  • NumPy
  • Pandas
  • OpenCV
  • Scikit-learn
  • PyTorch
  • Boost
CV background03

DevOps and automation

  • Git
  • GitLab CI/CD
  • Docker
  • Kubernetes
  • YAML
CV background04

Requirements and project tools

  • Jira
  • IBM DOORS
  • PTC Windchill
CV background05

Testing

  • Test automation
  • Unit testing
  • Integration testing
  • Model-based testing
  • Self-healing tests
  • Coverage analysis
  • Test plans
  • Test cases
  • Requirements traceability
Current academic work06

Current HPC and thesis tools

  • Linux
  • Slurm
  • vLLM
  • llama.cpp
  • LLM benchmarking
  • CPU performance analysis
  • Memory-bandwidth analysis
  • High-Bandwidth Memory

Open source and selected highlights

Work beyond the daily brief.

Selected contributions and projects from the CV, with live repository metadata remaining available on the Projects page.

Open source

Robot Framework

Active contributor to the Robot Framework open-source project, an automation framework for acceptance testing and robotic process automation.

Participated in Robot Framework RoboCon 2025 in Helsinki and represented ELT Group.

  • Robot Framework
  • Python
Open repository
Selected project

Jarvis: Personal Moodle Exam Assistant

A personal Moodle exam assistant built with browser automation and computer-vision tooling.

  • Python
  • Selenium
  • OpenCV
  • Scikit-learn
  • Pillow
Open repository
Selected project

Tiny-ViT

A computer-vision project exploring efficient Vision Transformers, including model pruning and quantization.

  • PyTorch
  • TensorFlow
  • Scikit-learn
Open repository

Languages and interests

Away from the terminal.

Running, riding, and photography offer a useful counterpoint to systems work and research.

Languages

Italian
Native
Romanian
Native
English
C1

Personal interests

  • Running
  • Cycling
  • Trekking
  • Nature photography
  • Urban photography

Full background

Experience, education, and technical work.

Open the synchronized PDF or download a copy for later.

Open CV

LDC / command line

Search and run a command

Type to filter. Use the up and down arrow keys to move between commands and Enter to run one.

Simulated portfolio diagnostic

Portfolio Diagnostics

A fictional, decorative sequence connecting software testing, HPC scheduling, LLM inference, and university research.

  1. Discovering requirements…Traceability matrix: ready
  2. Loading test architecture…Robot Framework libraries: ready
  3. Submitting compute workload…Scheduler: accepted
  4. Mapping model data…CPU ↔ HBM path: ready
  5. Collecting results…Portfolio verification: passed
Diagnostic ready to run.