Projects
Featured projects
Personal project
VIRA
An open-source AI-OS / Agentic OS layer that builds live context from your machine and verifies tools before they run.
Read More
Personal project
VIRA
An open-source AI-OS / Agentic OS layer that builds live context from your machine and verifies tools before they run.
VIRA is my open-source AI-OS / Agentic OS project. It is designed to keep a live model of system state and user behaviour, run fully locally, and synthesize new system apps, tools, and agents from natural language with verification before execution. The goal is to make the machine itself more context-aware without sending data to the cloud.
Repository: github.com/Vishalsng112/VIRA
SAC 2026
ODIN
A stochastic sampler for fault spectrums that supports statistically significant comparisons across spectrum-based fault localization metrics.
Read More
SAC 2026
ODIN
A stochastic sampler for fault spectrums that supports statistically significant comparisons across spectrum-based fault localization metrics.
Spectrum-based fault localization is effective and scalable, but the relative effectiveness of different metrics is often studied under limiting assumptions. ODIN samples fault spectrums from carefully selected control parameters so the resulting comparisons are more statistically grounded. Using ODIN, I studied 37 popular SBFL metrics and generalized several prior findings while also uncovering new equivalence relations that hold across experimental configurations.
Conference context: associated with fault localization research
ASE 2025
AndroFL
An infrastructure for evolutionary test generation and statistical fault localization for diagnosing Android app faults.
Read More
ASE 2025
AndroFL
An infrastructure for evolutionary test generation and statistical fault localization for diagnosing Android app faults.
AndroFL combines an evolutionary test-suite generator with a statistical fault localization engine. The test generator supports configurable fitness functions such as coverage and diagnosability metrics, while the localization module supports metrics including Ochiai, Tarantula, and Barinel. In our evaluation on 20 open-source F-Droid apps, the system reduced debugging effort substantially and localized more faults in the top ranked candidates than random testing.
Conference: 40th IEEE/ACM International Conference on Automated Software Engineering, Tool Demonstration Track
ASE 2023
Chiron
A teaching-oriented program analysis framework that packages analysis, verification, testing, and visualization into one modular system.
Read More
ASE 2023
Chiron
A teaching-oriented program analysis framework that packages analysis, verification, testing, and visualization into one modular system.
Chiron was designed to teach graduate-level program analysis, verification, and testing. It keeps the code base small, uses a unified intermediate representation, and offers a modular architecture for plugging in new algorithms. The framework includes dataflow analysis, abstract interpretation, symbolic execution, fuzzing, evolutionary test generation, and spectrum-based bug localization. It has also been used in multiple course offerings with very positive student feedback.
Conference: 38th IEEE/ACM International Conference on Automated Software Engineering, Research Track