HomeAboutProjectsContact
signal
signal
Sathwik Amburi
<Sathwik Amburi />
Sathwik Amburi avatarSathwik
About me
SA
Munich, Germany

Sathwik Amburi

PhD candidate at Siemens and Universität der Bundeswehr München researching how AI can improve secure software development in industrial settings.

My work sits between software security and AI engineering: designing useful context, building harnesses around model behavior, and turning research ideas into systems developers can actually use.

Security

AI for secure software development

Context

Task state and evidence design

AI Stack

Secure workflows and research prototypes

Education
  1. Education01
    April 2025 - Present
    education

    Doctor of Philosophy (PhD)

    Universität der Bundeswehr München, Germany

    AI for secure software development

    Doctoral research on leveraging AI for secure software development in industrial settings, with attention to context engineering, harness engineering, developer training, and evaluation.

    PhDSecure SoftwareContext EngineeringAI Engineering
  2. Education02
    October 2021 - February 2025
    education

    M.Sc. Informatics

    Technische Universität München (TUM), Germany

    AI privacy risk and secure software engineering

    Completed my Master of Science in Informatics with a focus on advanced software engineering, security, and AI. Co-authored my Master's thesis on 'A Systematic Privacy Risk Catalog for General Purpose AI Systems' at the Chair of Software Engineering for Business Information Systems (sebis).

    AI RiskSoftware EngineeringPrivacyTUM
  3. Education03
    July 2017 - May 2021
    education

    Bachelor of Technology in Computer Science and Engineering

    SRM Institute of Science and Technology, Chennai, India

    Computer science foundations and applied ML

    I completed my Bachelor of Technology in Computer Science and Engineering with a CGPA of 9.17/10. My studies included extensive coursework in software engineering, machine learning, and system design.

    Machine LearningSystemsAlgorithms
Experience
  1. Experience01
    April 2025 - Present
    experience

    Doctoral Researcher (PhD)

    Siemens AG & Universität der Bundeswehr München, Germany

    Leveraging AI for secure software development

    Researching how AI can support secure software development through better developer training, context-aware workflows, AI harnesses, and practical software-security tooling.

    Context EngineeringHarness EngineeringSecure CodingAI Engineering
  2. Experience02
    April 2023 - March 2025
    experience

    Working Student, Cyber Security

    Siemens AG, München, Germany

    Secure coding education, LLM evaluation, and research prototypes

    Created secure-coding challenges for developer education, evaluated LLMs for secure coding, analyzed Rust security vulnerabilities, and built research prototypes for GenAI-assisted secure development.

    Secure CodingRustLLM EvaluationResearch Prototypes
  3. Experience03
    January 2024 - September 2024
    experience

    Mentee, Siemens Mentoring Program

    Siemens AG, München, Germany

    Mentoring, innovation, and AR/VR prototyping

    Participated in the Siemens Mentoring Program 2024 and won the Siemens Mentoring Hackathon with an AR/VR solution for Hymer using TeamViewer Frontline Spatial Editor.

    MentoringHackathonAR/VRTeamViewer Frontline
  4. Experience04
    October 2023 - March 2024
    experience

    Frontend Developer (UI/UX)

    Parloa, München, Germany

    AI platform simulation and UI systems

    Developed a simulation model to streamline simulations, enhancing user interaction with the AI-driven platform. Led the design and implementation of user interfaces for Parloa's AI platform, focusing on dynamic interaction and testing of LLM bots.

    ReactLLM BotsUI/UX
  5. Experience05
    April 2023 - June 2023
    experience

    Software Developer

    Allianz GI, München, Germany

    Financial product interface and services integration

    Contributed to the development of an application that supports investors in meeting their investment goals. Developed a web interface to guide investors by integrating state-of-the-art risklab services.

    FrontendRisklabProduct
  6. Experience06
    October 2022 - March 2023
    experience

    Software Engineer, Interdisciplinary Project

    Forsailors, München, Germany

    Map-based route planning experience

    Created a visually appealing map style and route planner using Mapbox GL JS and React JS. Designed and implemented a route planner with relevant technologies.

    MapboxReactGeospatial
  7. Experience07
    March 2022 - March 2023
    experience

    Student Assistant (Programmer)

    TUM - Lehrstuhl für Baurealisierung und Baurobotik, München, Germany

    Computer vision tooling for building renovation

    Assisted in developing a Python plugin for FreeCAD that automatically creates accurate 3D building models from images. Contributed to computer-vision tooling for building renovation workflows.

    Computer VisionFreeCADBIM
  8. Experience08
    April 2020 - May 2020
    experience

    Intern

    MEIIPORUL SOLUTIONS PRIVATE LIMITED, Chennai, India

    Health analytics and epidemic modeling

    Assisted in the development of epidemic models like SEIR and SIR for forecasting COVID-19 spread. Designed a patient dashboard for monitoring health and developed deep learning models for categorizing chest X-rays and CT scans.

    Deep LearningDashboardsSEIR
Stack

Technical Proficiencies

Programming Languages
AI Engineering & Security
Frameworks & Interfaces
Databases
Engineering Practices
Cloud Platforms
Research

Publications

33Citations3h-index1i10-index
Google Scholar - updated 2026-06-21
Is This Mission Possible? A Study on Developer Challenges in Using Generative AI for Secure Software Development in Industry

Sathwik Amburi, Tiago Gasiba, Tobias Fertig, Ulrike Lechner, Maria Pinto-Albuquerque • ICT Systems Security and Privacy Protection (SEC 2026) • 2026

0 citations
GenAISecure Software DevelopmentSecure Code ReviewIndustryDeveloper Training

This chapter studies how industry developers use GenAI during secure software development and code review. Based on secure-coding workshops and survey responses, it identifies adoption patterns, perceived benefits, context limitations, hallucination risks, code-quality concerns, and practical strategies for responsible AI-assisted secure development.

View Publication
Enabling Secure Coding: Exploring GenAI for Developer Training and Education

Sathwik Amburi, Tiago Espinha Gasiba, Ulrike Lechner, Maria Pinto-Albuquerque • 6th International Computer Programming Education Conference (ICPEC 2025) • 2025

2 citations
Secure CodingGenerative AIDeveloper TrainingLLMsSoftware Security

This paper introduces an AI Secure Coding platform that embeds a GPT-4 based chatbot into a structured challenge workflow. It studies how GenAI can support developers as they identify, remediate, and reflect on software vulnerabilities, showing both productivity benefits and risks of over-reliance.

View Publication
Can Open Large Language Models Catch Vulnerabilities?

Diogo Gaspar Lopes, Tiago Espinha Gasiba, Sathwik Amburi, Maria Pinto-Albuquerque • 6th International Computer Programming Education Conference (ICPEC 2025) • 2025

0 citations
Large Language ModelsVulnerability DetectionCWE ClassificationSecure CodingCode Analysis

This paper evaluates open large language models on vulnerability detection and CWE classification using a filtered Big-Vul subset. The findings show strong detection behavior but weak fine-grained classification, highlighting limitations of current LLMs in security-sensitive workflows.

View Publication
Are We There Yet? On Security Vulnerabilities Produced by Open Source Generative AI Models and Its Implications for Security Education

Maria Camila Santos Galeano, Tiago Espinha Gasiba, Sathwik Amburi, Maria Pinto-Albuquerque • 6th International Computer Programming Education Conference (ICPEC 2025) • 2025

0 citations
Generative AICode SecurityProgramming EducationPrompt EngineeringStatic Analysis

This paper studies security vulnerabilities produced by open source code generation models across Python, C, and Java prompts. It identifies recurring flaws such as command execution issues, insecure memory handling, and weak input validation, then proposes training practices for safer GenAI-assisted development.

View Publication
May the Source Be with You: On ChatGPT, Cybersecurity, and Secure Coding

Tiago Espinha Gasiba, Andrei-Cristian Iosif, Ibrahim Kessba, Sathwik Amburi, Ulrike Lechner, Maria Pinto-Albuquerque • Information • 2024

19 citations
EducationTrainingSecure codingIndustryCybersecurityCapture the flagGame analysisCybersecurity Challenges

This paper explores how ChatGPT can aid secure software development, drawing on experiments with large language models and prior secure coding research. It discusses advantages, limitations, and risks for code analysis, developer guidance, and cybersecurity education.

View Publication
Can Secure Software be Developed in Rust? On Vulnerabilities and Secure Coding Guidelines

Tiago Espinha Gasiba, Sathwik Amburi, Andrei-Cristian Iosif • International Journal On Advances in Security • 2024

1 citation
Rust ProgrammingSecure CodingSoftware Security

The paper examines the security of software developed in the Rust programming language, comparing it to C, C++, and Java. It highlights Rust's strengths in memory safety and concurrency but notes that writing secure software is still challenging. The study identifies ten common security pitfalls in Rust and suggests that while Rust improves security, vulnerabilities are still possible. The findings are based on literature reviews and expert interviews, contributing valuable insights for both academia and industry.

View Publication
Online Modelling and Prefab Layout Definition for Building Renovation

Kepa Iturralde, Sathwik Amburi, Sandhanakrishnan Ravichandran, Samanti Das, Danya Liu, Thomas Bock • International Symposium on Automation and Robotics in Construction (ISARC 2023) • 2023

5 citations
Building ModelRenovationPrefabricationAutomationBIM

This paper introduces a semi-automated tool for creating detailed 3D building models and prefabricated module layouts from building images and OpenStreetMap floor plans. The workflow supports faster renovation planning and analysis of solar-panel placement on building envelopes.

View Publication
I Think This is the Beginning of a Beautiful Friendship - On the Rust Programming Language and Secure Software Development in the Industry

Tiago Espinha Gasiba, Sathwik Amburi • CYBER 2023 • 2023

4 citations
Rust ProgrammingSecure CodingSoftware Security

The paper examines the Rust programming language, highlighting its security benefits and comparing it to C, C++, and Java. It discusses Rust's strengths, like memory safety, while acknowledging that vulnerabilities can still exist. The study uses literature review, expert interviews, and static analysis tools to assess Rust's security, contributing to both academic knowledge and practical industry insights. Despite its advantages, the research emphasizes that Rust isn't immune to security issues and encourages ongoing vigilance and further research.

View Publication
Transforming Native Epidemic Models by Using the Machine Learning Approach

Sathwik Amburi, Vanshika Jalan, Dr. S. Saravanan • Annals of the Romanian Society for Cell Biology • 2021

2 citations
Epidemic ModelsDeep LearningSEIRD Model

This paper explores transforming traditional epidemic models using machine learning. Conventional models often fail to adapt to real-world complexities and evolving socio-environmental factors. By integrating machine learning, specifically deep learning techniques, the paper proposes more dynamic models that can better forecast and respond to epidemic conditions, such as those seen during the COVID-19 pandemic. The approach includes advanced data collection, preprocessing, and visualization to improve predictions and inform public health strategies.

View Publication