Timebox: ~60 minutes. Focus on high-impact changes and clear reasoning.
Choose 1 task that fits your role/experience best:
agent.py – Fix/upgrade the reasoning/guardarail capabilities of this agent.
model_finetuning.py – Upgrade the finituning strategy for this model.
rag.py – Correct and improve robustness of a RAG pipeline.
The goal of this task is not to see how well you can use AI. All scripted were based on AI generated code, so we are looking for how you can break down a system and improve it. We run an SOTA agent over it and have a list of easy improvements. Instead, focus on the abstract ideas: improvements on the used methods, system design, or clear algorithmic optimizations.
We are not looking for a huge list of changes. Choose 1-2 clear improvements that demonstrate your coding/algorithmic/ML capabilities.
-
Click Use this template on the repo page → create a Private repo in your account. Do not fork; forks may be visible to others.
-
Do the work (≤60 min).
-
Make a commit (or optional PR) that includes:
-
delivery.md(your report; see template) -
Any code changes you made (optional but welcome)
- Add reviewers as collaborators with repo access:
joaocarv, welyweah, viictorjimenezzz, gavi210
- Share the repo (or PR inside your private repo) link with us.