prats.page

Master's in Informatik – Coursework & Technical Archive

| 3 min read

This document is a structured academic record of the courses I completed during my Master’s degree at RPTU Kaiserslautern and the key systems, concepts, and implementations I worked on.
It serves as a comprehensive archive of my coursework.

Academic Timeline

Winter Semester 2023/24 (1st Semester)

  • Network Security
  • Protocols and Algorithms for Network Security
  • Database Systems
  • Very Deep Learning – Recent Methods and Technologies
  • Social Web Mining

Summer Semester 2024 (2nd Semester)

  • Machine Learning I
  • Verification of Reactive Systems
  • Performance and Security Analysis

Winter Semester 2024/25 (3rd Semester)

  • Decentralized Systems
  • Worst-Case Analysis of Distributed Systems
  • OS-Based Programming of Embedded Systems
  • Seminar: Privacy and Security
  • Programming Distributed Systems

Summer Semester 2025 (4th Semester)

  • Distributed Data Management
  • Human-Computer Interaction
  • Practical Machine Learning & NLP with AWS
  • Engineering with Generative AI
  • Project: Secure Decentralized Systems

Distributed Systems & Infrastructure

Programming Distributed Systems

Project: MiniDote – Distributed Key-Value Store
Repository: https://github.com/lawRathod/minidote

Covered Topics

  • RPC / gRPC-based inter-node communication
  • Replication and synchronization mechanisms
  • Crash failure models
  • Network partition simulation
  • Consistency vs availability trade-offs
  • Distributed testing methodologies

What I Implemented

  • A replicated distributed key-value store
  • Inter-node synchronization logic
  • Failure scenario simulations
  • Convergence and correctness evaluation under partial failures

Distributed Data Management

Covered Topics

  • Data sharding and partitioning strategies
  • Replication models
  • Distributed transactions
  • Storage consistency models
  • Trade-offs between performance and correctness

Worst-Case Analysis of Distributed Systems

Covered Topics

  • Adversarial and worst-case system modeling
  • Latency bounds under failures
  • Formal reasoning about distributed correctness
  • Resilience analysis under constrained system assumptions

Decentralized & Secure Systems

Decentralized Systems

Project: IPFS Lab & Kubo Modification
Repository: https://github.com/lawRathod/ipfs-lab

Covered Topics

  • Peer-to-peer networking architectures
  • Content-addressable distributed storage
  • Peer discovery and bootstrapping mechanisms
  • Distributed topology formation
  • Network convergence behavior

Additional Work

  • Patched Kubo (IPFS implementation) to disable automatic bootstrapping
  • Conducted controlled peer topology experiments
  • Evaluated replication and connectivity dynamics
  • Observed network behavior under constrained bootstrap conditions

Project: Secure Decentralized Systems

Repository: https://github.com/lawRathod/uni-project-dec

Covered Topics

  • Threat modeling for distributed systems
  • Secure multi-node architecture design
  • Authentication and communication integrity
  • Adversarial resilience evaluation
  • Secure protocol integration

Network Security

Covered Topics

  • TLS and secure channel establishment
  • Authentication mechanisms
  • Secure API design principles
  • Network threat modeling

Protocols and Algorithms for Network Security

Covered Topics

  • Cryptographic protocol construction
  • Secure key exchange mechanisms
  • Protocol-level security analysis
  • Formal reasoning about secure communication
  • Security-performance trade-offs

Seminar: Privacy and Security

Seminar Focus

Asynchronous Vertical Federated Learning (AVFL)

Covered Topics

  • Federated learning paradigms (horizontal vs vertical)
  • Vertically partitioned collaborative training
  • Asynchronous model update mechanisms
  • Delayed gradient challenges
  • Resource heterogeneity in distributed learning systems
  • Communication constraints in federated environments

Methods Studied

  • Asynchronous Stochastic Gradient Descent (ASGD)
  • Variance reduction techniques (SVRG, SAGA)
  • Gradient prediction via Taylor approximations
  • Secure aggregation protocols
  • Cryptographic privacy-preserving mechanisms

Key Observations

  • Trade-offs between efficiency, accuracy, and privacy
  • Convergence behavior under asynchronous updates
  • Communication-efficient optimization strategies
  • Resource utilization improvements over synchronous approaches
  • Architectural considerations for scalable privacy-preserving learning

Systems Performance & Verification

Performance and Security Analysis

Covered Topics

  • Latency and throughput benchmarking
  • Profiling distributed systems
  • Bottleneck identification
  • Performance tuning methodologies
  • Security-performance interaction analysis

Verification of Reactive Systems

Covered Topics

  • Model checking
  • State machine verification
  • Formal correctness guarantees
  • Specification validation of reactive systems

OS-Based Programming of Embedded Systems

Covered Topics

  • Concurrency primitives
  • Resource-constrained systems programming
  • Low-level memory and hardware interaction
  • Real-time execution considerations

Machine Learning & AI Systems

Machine Learning I

Covered Topics

  • Supervised learning algorithms
  • Model evaluation techniques
  • Regularization and optimization
  • Bias-variance trade-offs
  • Generalization analysis

Very Deep Learning – Recent Methods and Technologies

Covered Topics

  • Transformer architectures
  • Deep neural optimization strategies
  • Scaling considerations for large models
  • Compute and memory trade-offs
  • Empirical evaluation of architectures

Practical Machine Learning & NLP with AWS

Covered Topics

  • Cloud-based ML deployment
  • Model serving architectures
  • CI/CD pipelines for ML workflows
  • Monitoring and scaling ML systems
  • Infrastructure considerations for production ML

Engineering with Generative AI

Project: Generative AI Portfolio
Repository: https://github.com/lawRathod/gen-ai-portfolio

Covered Topics

  • LLM orchestration patterns
  • Prompt engineering strategies
  • Structured output generation
  • Inference pipeline design
  • Latency-aware model integration

What I Built

  • Modular LLM-based applications
  • Inference workflows integrating external APIs
  • Evaluation mechanisms for output reliability

Social Web Mining

Covered Topics

  • Large-scale data collection methods
  • Feature extraction pipelines
  • Data preprocessing and cleaning
  • Statistical analysis of heterogeneous and noisy datasets

Databases & Human-Centered Systems

Database Systems

Covered Topics

  • Indexing strategies
  • Query optimization
  • Storage engine fundamentals
  • Transaction isolation levels
  • Database performance considerations

Human-Computer Interaction

Covered Topics

  • Usability evaluation methods
  • Interface design principles
  • User-centered system design
  • Interaction modeling and evaluation techniques

Core Themes Across the Degree

  • Distributed systems implementation
  • Replication and synchronization
  • Secure multi-node architectures
  • Performance benchmarking and tuning
  • Fault tolerance modeling
  • Privacy-preserving distributed learning
  • Cloud-based ML deployment
  • Deep learning and generative AI systems
  • Protocol-level security reasoning
  • Low-level systems programming