Autonomous optimization for code and algorithms.

Need code optimized? Get in touch.

Optimize my code ↗View the workflow ↗

✓ Original preserved · Outputs checked · Failed attempts recorded

REPOSITORY-FIRST·MEASUREMENT-DRIVEN·AI-AGNOSTIC·OPEN SOURCE
Examples

Real code. Repeated attempts.
Visible improvement.

GRID ROUTE

Finding the same route with fewer checks

Same shortest route.

99.8% faster27.664 ms → 0.0517 ms · 1,200 trials
APPLE CHIP CALCULATION

Reducing a repeated calculation

Same answer.

54.5% faster1.148 ms → 0.522 ms · 40 A/B trials
MATRIX CALCULATION

Rearranging work for better memory use

Same matrix result.

92.3% faster3.904 ms → 0.302 ms · 30 trials
DATA LOOKUP

Adding a shortcut for repeated row searches

Same rows returned.

70.0% faster2.517 ms → 0.754 ms · 30 trials
SCROLL SIDEWAYS FOR MORE
Request an optimization

Need code optimized?

Send the code, benchmark, or bottleneck. I’ll reply by email.

Send a request

Submissions go directly to Vuk’s inbox.

Open email ↗X ↗LinkedIn ↗
Run it yourself

Install the optimization skill.

COPY-PASTE PROMPT
Install and use the `auto-optimize` skill from this repository:

https://github.com/vukrosic/open-discovery/blob/main/.agents/skills/auto-optimize/SKILL.md

Read it and start.
How it works

One task. Measured result.

01

Freeze the baseline

02

Run candidates

03

Check outputs

04

Keep the evidence

Looking for a cofounder

Build the next layer of scientific progress.

Open Discovery starts with algorithm optimization and aims toward automated digital research. Read the full message if you want to build it together.

Read the message ↗