The 2024 Term 3 of the LFX Mentorship program is here, and it's packed with exciting opportunities! Running from September to November, this program invites passionate developers to contribute to open source while boosting their CV. WasmEdge has 4 projects that offer an exciting opportunity for aspiring developers to work on cutting-edge projects within the WasmEdge ecosystem, with a focus on enhancing WebAssembly (WASM) capabilities, improving software reliability, and integrating modern AI techniques to build innovative applications. Let's explore the projects and what mentees can expect.
Be mindful about the application deadlines. If you have questions, make sure to only contact the team and mentors publicly (to be fair to all candidates, private emails or messages will not be responded to as stated in this announcement). If you’d like to learn more about the LFX mentorship, check out the last section of the WasmEdge August monthly community meeting.
WASM Serializer with New Proposals
Project Description:
WasmEdge provides a WASM module serializer at the C API level, allowing developers to convert loaded WASM structures back into binary format. With recent support for function-references, GC, relaxed-SIMD, and exception-handling proposals, the serializer needs updates to handle these new partitions. Mentees will work on completing the binary format serialization for these new WASM proposals.
Expected Outcome:
- Complete the serialization of the new module extensions in WASM proposals.
- Complete the serialization of the new instructions added in WASM proposals.
- Add some basic unit tests with hand-writing WASM binaries.
Recommended Skills:
- C++, WASM, git
Upstream Issue: WasmEdge Issue #3585
LFX URL: WASM Serializer Project
Fix Bugs Found by Fuzzer
Project Description:
WasmEdge has received several bug reports identified by Fuzzer. Mentees will investigate these reports to determine their validity, propose solutions, or mark issues as won't-fix if invalid. Applicants must submit a proposal as part of their application.
Expected Outcome:
- At least fix/determine 60% of the mentioned issues.
Recommended Skills:
- git, C++, WebAssembly
Upstream Issue: WasmEdge Issue #3584
LFX URL: Fuzzer Bug Fix Project
Create an LLM App with Deep Understanding of a GitHub Repo
Project Description: Leveraging LLM (Large Language Models) for coding assistance can significantly impact open-source development. This project aims to build LLM agents using LlamaEdge and WasmEdge, supplemented with deep knowledge of open-source projects on GitHub. The goal is to enable the agent to answer questions and solve problems raised by the community.
Expected Outcome:
- Build an automated tool to extract and process all files in a repo. That includes source code and docs.
- develop a GitHub bot to capture all change files and update the knowledge base in real time.
- generate a summary for each file (using an LLM) and supplement with its file path and other meta data.
- create a vector database with the summary and original text. The vector is computed from the summary to improve search efficiency.
- Run an LLM agent node with the RAG database from the repo.
- Create a GitHub bot that can read new issues and respond with either an answer or a coding suggestion based on the content inside the repo.
- Evaluate the answer quality.
Recommended Skills:
- Rust, LlamaEdge, ChatGPT and LLMs, RAG process
Upstream Issue: WasmEdge Issue #3581
LFX URL: LLM GitHub Repo Project
Create a Wasm-Based LLM App for Financial Analysts
Project Description: This project involves developing an LLM-based financial data analytics application using open-source LLMs, embedding models, the LlamaEdge application server, vector databases, and data processing tools. The aim is to create a “template” and showcase “best practices” for similar applications in the financial sector.
Expected Outcome:
- Create a data processing pipeline in Python or Rust to automatically
- collect public company’s SEC 10-Q quarterly reports and press releases. e.g., Apple 10-Q and Apple press release
- generate a summary for each SEC 10-Q and press release documents using an LLM service such as LlamaParse or EYELEVEL xRay
- create and continuously update a vector database with the summary and original text. The vector is computed from the summary to improve search efficiency.
- Create a server-side RAG app that can chat with the vector knowledge base of financial statements.
- Evaluate the answer quality
- Explore LLM function calling to incorporate real-time information and actions
Recommended Skills:
- Python, LlamaEdge, ChatGPT and LLMs, RAG process, Rust (optional)
Upstream Issue: WasmEdge Issue #3580
LFX URL: LLM Financial App Project
Join the Future of WebAssembly and AI with WasmEdge
The LFX Mentorship program provides a unique opportunity for mentees to work on significant projects with real-world impact. Whether you're looking to sharpen your skills or interested in WebAssembly, or AI-driven applications, these projects a fantastic way to learn while gaining invaluable experience. Apply now to join us in this journey of innovation, where you'll work alongside experienced mentors, tackle real-world problems, and help shape the future of technology. Let's build something amazing together!