Find open source projects actively accepting contributors. Search repositories, filter by program milestones, difficulty tags, or tech stack.
Use our Orbit AI Matcher to find out! Get instant matching scores based on your developer skills, preferred frameworks, and contribution experience.
Convert your selected open-source project into a winning GSoC, LFX, or Outreachy application using Proposal Studio.
This project aims to develop an AI-powered system that supports early detection and ongoing monitoring of suicide, substance use, and mental health crises through social media analysis. By combining behavioral tracking, natural language processing (NLP), and geospatial trend mapping, the system will extract high-risk language patterns, identify regions of escalating distress, and visualize crisis trends over time. The final deliverables will include a structured crisis dataset, predictive risk indicators, and an interactive dashboard to help clinicians, nonprofits, and public health agencies make timely and informed decisions. This project addresses the limitations of delayed traditional health data by providing real-time insights into emerging community crises.
This project implements a comprehensive GitHub-Discord integration platform. The system includes a Discord bot with OAuth GitHub account linking for automatic role assignment, a PR review automation system with AI-powered code analysis, and a data pipeline that processes and stores GitHub events for analytics. Built on Python with Flask and Discord.py, deployed via Google Cloud Run, and orchestrated through GitHub Actions workflows.
As interest in large language models continues to grow, many open-source implementations exist across various frameworks—but accessible, modular, and well-documented implementations in Flax, built on JAX, remain limited. This lack of resources makes it more difficult for developers and researchers to understand, experiment with, and extend LLMs in the JAX ecosystem. This project addresses that gap by implementing one or more open LLM architectures in Flax. The goal is to provide reference models that are easy to read, modify, and build upon. The project will also include a comprehensive test suite to ensure reliability, and an interactive educational notebook that explains the architecture, shows how to load pretrained weights, and demonstrates how to run inference.
This project extends the open-source Cormas simulation platform by integrating a computer vision system that allows real-world interaction with agent-based models. The goal is to enable users to turn Cormas simulations into interactive, game-like experiences by detecting and tracking physical objects (such as game pieces) on a board and linking them to simulation agents. To achieve this, I will develop a lightweight Python-based vision pipeline using YOLO for object detection and tracking, then build a JSON-based interface to communicate with Cormas in real time. On the Pharo/Cormas side, I will implement the necessary APIs and tools to interpret the visual data and control agent behavior accordingly. Deliverables include: - A reusable computer vision module with manual calibration tools. - A JSON-based interface for real-time communication between Python and Pharo. - Integration into Cormas as a plugin or extension. - Documentation and examples for creating interactive physical games with Cormas. - All work will be contributed to the open-source Cormas repository and designed to support future projects involving physical interaction with simulations.
AEGIS automates hardware verification by using AI to generate optimized test cases. Parses Specifications: Extracts key features (interfaces, protocols) from documents using NLP. Generates Tests: Creates corner-case and compliance tests via ML models. Optimizes Prioritization: Uses RL to maximize coverage efficiency. Outputs UVM Code: Auto-generates UVM testbenches for simulation.
This project develops cloud-native tools to detect novel insecticide resistance in African malaria mosquitoes. Current PLINK-based methods are insufficient as VGSC L995F mutation is now fixed and new resistance mechanisms have undefined targets. The solution implements advanced statistical methods (mixed-effects models, Bayesian approaches), population structure correction, interactive visualization tools, and cloud optimization for accessibility. Following a tiered strategy, core deliverables include statistical methods and visualization, with machine learning extensions if time permits. Leveraging my interdisciplinary background, this project will improve resistance detection sensitivity, helping researchers identify emerging resistance mechanisms before they become widespread, contributing to more effective malaria vector control.
The VCell-AI Platform is an AI-powered system for discovering, analyzing, and exploring biomodels, publications, and research papers from the VCell database. It supports both traditional database queries and AI-powered natural language searches through a chatbot, provides biomodel diagram analysis, and includes a vector-based knowledge base for document retrieval for VCell Software tutorials and guides. Built with Next.js and Tailwind CSS on the frontend, FastAPI and Python on the backend, and Qdrant for vector storage, the platform also integrates Langfuse for LLM observability and can run with locally hosted LLMs, offering researchers a fully functional, flexible, and user-friendly environment for scientific exploration.
The LiftOver2D project aims to develop a high-performance tool for converting Hi-C matrices between different genome assemblies and binning schemes. This functionality is essential for integrating data from various sources and experiments — especially with the growing popularity of multi-omics approaches and the increasing availability of public datasets. Existing solutions, such as HiCLift, offer partial support but often fail to provide an optimal balance between accuracy and performance, particularly in cases of partial bin mismatches or complex remapping scenarios. LiftOver2D seeks to overcome these limitations by leveraging background contact distributions at the bin level, allowing for more precise and efficient remapping. The main goal is to design a method that achieves an acceptable trade-off between accuracy and speed, ensuring the tool is practical for real-world bioinformatics pipelines without sacrificing data quality or runtime efficiency. During the project, we plan to develop a custom model for bin-level remapping and optimal bin-to-bin matching across genome assemblies, and test our tool on various experiments and organisms using publicly available data from open databases (such as NCBI GEO). We aim to maximize performance and parallelization using both high-level data frameworks (like Dask or Polars) and lower-level languages where appropriate. The resulting algorithm will be wrapped into convenient packages for Python, R, and the command line. We also plan to collaborate with major tools in the Hi-C ecosystem (such as Cooler or Muon for multimodal experiments) to integrate our functionality into widely-used frameworks. In the long term, LiftOver2D aspires to become a community-standard solution for fast remapping of 2D genomic contact data. The development will be carried out as an open-source project in collaboration with the community and aligned with NumFOCUS’s mission to support sustainable scientific software.
This project will focus on improving Zulip web app and shipping new features. Deliverables focuses on zulip message system and permission settings but this will probably end up reaching wider areas of Zulip.
Information Delivery Specification (IDS) is a standard for defining information requirements in a computer interpretable form, written in XML. Manually writing XML in compliance with the IDS schema can be an extremely difficult task for non-technical BIM users, giving rise to a need for a web interface to compose IDS specifications in a user-friendly manner. The current implementation suffers from bugs, outdated UI, and lacks important authoring usability features. The primary goal of this proposal is to rebuild the web app on a new foundation - the IfcOpenShell WASM build, eliminating the need for a API backend, improving privacy, and enabling innovative features like in-browser IFC compliance checking, contextual autocompletions, and enhanced UX/UI. The new implementation will be comprised of 3 modules: 1) IDS Module - for composing and editing IDS specifications in a user-friendly way. 2) WASM Module - to interface with IfcOpenShell WASM. 3) Application Module - built with Svelte 5 and shadcn-ui to deliver a modern, responsive UI. Additionally, this proposal introduces a built-in XML (IDS) editor, in-browser audit report generation, and integration with Bonsai’s websocket-based Web UI system for fetching "smart context" from Blender.
The project Scrum Helper is designed to aid individuals in generating and injecting daily scrums into popular email clients such as Gmail, Outlook, and Yahoo. This initiative focuses on automating repetitive tasks involved in daily scrum reporting by leveraging user activity data from platforms like GitHub. The goal is to streamline scrum generation, improve developer productivity, and promote efficient communication within teams. Key Features: 1.Cross-Platform Support: Generate scrums in Gmail, Outlook, Yahoo, or via a standalone popup. 2.GitHub Integration: Auto-fetch activity to pre-fill scrum content. 3.Customizable & Shareable: Preview, edit, copy, and save scrum templates. 4.Secure & Optimized: OAuth-based access, responsive UI, and error handling.
Virtio-fs is a shared file system that lets virtual machines access a directory tree on the host. Virtiofsd is the user-level host daemon that communicates with the guest VMs and performs I/O operations on the shared directory tree. Currently, virtiofsd processes FUSE-based I/O requests in a mostly sequential manner. Even though it uses a thread pool, concurrency is limited by the number of available threads. This project aims to enable truly asynchronous I/O request handling in virtiofsd by leveraging io_uring. The deliverables will include a complete patch implementing asynchronous I/O handling, along with a performance evaluation of the resulting system.
This project aims to add DVD-Audio support to VLC by implementing a module that decrypts and plays back CPPM-protected audio streams. Using libdvdcpxm and related tools, the module will extract, decrypt, and decode high-fidelity DVD-Audio, ensuring seamless integration with VLC's existing playback architecture. The final deliverables include a working VLC module, full documentation, and optimizations for performance and compatibility.
ROCm is one of the most promising open-source platforms for GPU computing, providing a vendor-neutral alternative to CUDA for scientific, AI, and HPC workloads. Although Debian ships with some of its core components, GPU acceleration is still underutilized in many high-impact packages. This project addresses that by identifying such packages where ROCm support is available upstream, enabling GPU acceleration in those packages, contributing new policy-compliant Debian packages that will benefit the ROCm ecosystem and ensuring correct functionality via autopkgtest integration.
In high-energy physics experiments such as those conducted at CERN's ATLAS project, the volume of data generated is immense, reaching up to 60 million MB per second. While lossless compression is already employed to manage this data, lossy compression—specifically of floating-point precision—offers more aggressive reductions, potentially decreasing file sizes by over 30%. However, this comes at the cost of irreversibly discarding information, raising the challenge of how to recover or approximate full-precision data for downstream analysis. This project proposes a novel approach using deep probabilistic models to reconstruct high-precision floating-point data from aggressively compressed representations, a problem coined as precision upsampling. The goal is to explore and compare the capabilities of three classes of generative models: autoencoders, diffusion models, and normalizing flows. Each model type offers distinct advantages: autoencoders are well-studied in neural compression, diffusion models are robust to noise and excel in reconstructing multi-scale structures, and normalizing flows offer exact likelihood estimation and invertible mappings that align well with the structured nature of physical data. Deliverables include a literature review on neural precision upsampling methods, a working baseline model adapted to CERN's PHYSLITE floating-point compressed data, and a comparative study of advanced probabilistic models—including variational autoencoders, diffusion models, and normalizing flows. Evaluation metrics will also be tailored to the data and implemented. Final outputs will include reproducible code, performance benchmarks, a written technical report, and optionally a publication or integration into the ATLAS codebase.
The IntelliJ Platform Gradle Plugin, the official Gradle plugin for building, testing, verifying, configuring environments, and publishing plugins for IntelliJ IDEs, includes a VerifyPluginTask that verifies a plugin against one or multiple IDEs. Currently, reporting is managed by the Verifier CLI, which employs its custom reporting system to generate verification reports; however, the VerifyPluginTask could benefit from Gradle's Reporting API for more configurable reports. Additionally, integrating Gradle's Worker API for parallel execution of the verifyPlugin task could enhance the IDE plugin development experience. The goal of this project is to assess the benefits and ultimately implement the necessary changes to the verifyPlugin task. Lastly, explore further optimizations to enhance the plugin development workflow.
This project aims to enhance the user experience for both challenge hosts and participants by developing an intelligent, RAG (Retrieval Augmented Generation) based chatbot. The chatbot will efficiently address queries related to challenge hosting, guidelines, troubleshooting, and FAQs. By integrating state-of-the-art NLP techniques with robust retrieval mechanisms, the solution will ensure prompt, accurate, and context-aware responses that reduce support overhead and streamline communication. Using the RAG approach, the chatbot will retrieve relevant information from challenge documentation and combine it with generative models to create coherent and helpful answers. This will empower hosts to manage challenges more effectively and assist participants in resolving queries, ultimately contributing to a smoother and more interactive challenge experience. Deliverables include : RAG Framework Implementation Chat UI Comprehensive Documentation Testing, Feedback, and Continuous Improvement
This GSoC proposal aims to create an Assets Library for MIT App Inventor, addressing the challenge of managing assets (images and sounds) across multiple projects. The solution involves developing a responsive Designer panel, distinct from the AssetListBox, to upload, organize, and import assets. I’ll leverage the UserFileData structure for user-specific storage, integrate with Google Cloud Storage (GCS), and refactor existing UploadServlet logic for robust validation. Deliverables include a drag-and-drop UI, folder-based organization, batch import functionality, and UX enhancements like search and error handling. This 175-hour project, spanning June 2 to August 25, 2025, will enhance workflow efficiency for educators, students, and developers, aligning with App Inventor’s mission to simplify app creation.
The multibeam sonar is a ray-based plugin designed to simulate the physics of acoustic signal propagation, including phase and reverberation effects. Originally developed for ROS 1 Noetic and Gazebo Harmonic as part of the DAVE project, it is currently being migrated to Gazebo Harmonic and ROS 2 Jazzy. This project aims to complete the migration and enhancement of the multibeam sonar plugin, with three main deliverables. The first is an enhanced plugin with improved performance through the introduction of half-precision calculations in CUDA. It is expected that this modification will greatly improve frame refresh rates. Numerical accuracy must be carefully taken into account when implementing it, however, and this should involve thorough testing and evaluation using quantitative metrics in comparison to the performance of the single-precision code. The second deliverable is the extension of simulation capabilities to support additional sonar types. Building on the existing multibeam sonar, it should be feasible to implement a side-scan sonar with for seafloor imaging, taking into account shadow analysis and directivity patterns like side lobes, for example. Finally, benchmark integration is essential to make the simulation more practical and accessible. This involves incorporating hardware that is often used in maritime robotics, like the BlueROV, and not only efficiently modeling the vehicle and its sensors, but also offering examples that are effortless to use. These benchmarks will enable the simulation to be validated and more available and beneficial for the general community, researchers, and developers.
This project extends Dagger.jl with an MPI-based backend, allowing it to scale more effectively on high-performance computing (HPC) clusters that use low-latency interconnects like Infiniband or HPE Slingshot. Dagger already offers a productive way to express parallelism in Julia, using Directed Acyclic Graphs (DAGs) to schedule and run tasks. However, it currently relies on Julia’s Distributed.jl and TCP for communication, which becomes a bottleneck at large scales. In the exaflop era, where extreme-scale simulations demand both scalability and productivity, this limitation is significant. By integrating MPI, we build on vendor-optimized communication libraries to combine Dagger’s ease of use with the performance needed in HPC environments, much like frameworks such as Legion and Iris.
This project aims to develop a Text-based User Interface (TUI) framework for Dart, addressing the lack of a structured way to build interactive terminal applications in the language. While Dart excels in GUI development with Flutter, it lacks a robust solution for TUI-based applications. This framework will provide essential UI components, efficient rendering, and event-driven input handling, enabling developers to create modern, cross-platform terminal applications. Key deliverables include a modular TUI library with core UI components, theming support, optimized rendering, comprehensive documentation, and real-world examples. This project will expand Dart’s capabilities beyond mobile and web, making it a viable choice for terminal-based development.
The goal of this project is to add support for third-party themes to customize tab group colors in Chromium. Currently, users are limited to a fixed set of predefined tab group colors. This project will enable theme authors to define their own color palettes, offering greater flexibility and personalization. To achieve this, I will design an API to support customizable tab group colors, implement validation logic to ensure the correctness of user-defined colors, and generate shades dynamically from a single hue using a state-of-the-art algorithm. These colors will be seamlessly integrated into the browser’s color pipeline. Additionally, I will update the theme storage and retrieval mechanisms to support disk persistence. The key deliverables include the API extension, comprehensive tests, and documentation. This feature will greatly enhance how users organize tabs and boost productivity.
The Entity Display Manager module aims to simplify how entity displays are handled in Drupal. Currently, managing view modes and form displays across different entities can be tedious and repetitive. This project will introduce a centralized, user-friendly interface to configure, duplicate, and synchronize entity displays efficiently. By providing intuitive tools and supporting bulk operations, the module will improve developer productivity and site-building experience in Drupal.
Machine learning based analyses are proving to be very potent in understanding complex physical systems. This project aims to develop a reinforcement learning - CNN hybrid machinery that will perform unsupervised super-resolution and study of observed strong gravitational lensing images. The developed machinery will be useful in the study of data releases of upcoming lensing surveys.