Building high-performance development workflows means choosing the right tools — from blazing-fast package managers to AI agents that can reason, refactor, and rewrite code.
At IndyPy’s April 2025 meetup, attendees got a front-row seat to two live demos:
David Birks (E-gineering consultant) showcased how uv, a Python package manager, is setting speed benchmarks and simplifying virtual environments.
Calvin Hendryx-Parker (Six Feet Up CTO) ran a live “Battle of the Bots,” testing five popular AI coding assistants — Aider, Claude Code, Cursor, Goose and Junie — on a real-world dev task.
uv: A Smarter Python Package Manager
David Birks demonstrated how uv, a pip-compatible package manager written in Rust, is raising the bar for Python tooling. Designed to streamline dependency management and eliminate long install times, uv consolidates several common tools into one efficient solution.
Highlights:
Replaces pip, poetry, pyenv, and venv with a single unified workflow
Caches dependencies globally to speed up subsequent installs
Handles Python version management automatically
Reduced install time from 12 minutes to 90 seconds on a constrained Windows machine
For teams juggling virtual environments and complex dependencies, uv offers a dramatically faster and more modern developer experience.
Battle of the Bots: AI Assistants Showdown
Calvin Hendryx-Parker compared five popular AI development tools: Aider, Claude Code, Cursor, Goose, and Junie. Each assistant was given the same prompt — convert a Django project’s GraphQL API to REST — using a codebase scaffolded with Scaf™.
The results showed differences in performance, reasoning, and reliability:
Claude Code performed the most complete refactor, modifying dozens of files with minimal prompting and high accuracy.
Aider offered fine-grained Git control and multi-model customization, but needed more coaching to complete tasks.
Cursor shined with its context-aware “rules” engine and VS Code integration — especially helpful for team standards.
Goose introduced session-based workflows and strong tool access via the MCP protocol but suffered reliability hiccups during the demo.
Juni, JetBrains’ AI assistant, showed promise but felt slower and more rigid compared to its peers.
“These aren’t just autocomplete tools,” Calvin said. “They’re decision-makers — and how well they understand your architecture really matters.”
Speed isn’t optional: Tools like uv and Claude Code reduce friction and save hours
AI assistants aren’t interchangeable: Choose based on your stack, team needs, and workflow
Project conventions improve results: Assistants perform better when guided with structure
Cost matters: High-powered models can burn through tokens quickly
Agentic tools require setup: Protocols like MCP unlock power, but need thoughtful integration
Whether you're optimizing a dev team, evaluating AI pair programming, or modernizing your Python workflows, the right tools can make all the difference.
Ready to explore what AI, automation, and scalable Python systems can do for you? Let’s talk.
Watch the Full Presentation
Calvin’s“Battle of the Bots” begins at the 21:38 mark.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
