# Package technical spec feedback

**URL:** <https://discourse.julialang.org/t/package-technical-spec-feedback/139056>\
**Category:** Data\
**Tags:** package, spec-driven-dev, synthetic-data\
**Created:** [August 26, 2026, 5:14pm UTC](https://discourse.julialang.org/t/package-technical-spec-feedback/139056 "2026-08-26T17:14:35Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![mthelm85](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mthelm85/32/224164_2.png) [@mthelm85](https://discourse.julialang.org/u/mthelm85)\
**Post date:** [August 26, 2026, 5:14pm UTC](https://discourse.julialang.org/t/package-technical-spec-feedback/139056/1 "2026-08-26T17:14:35Z")

</div>

About 6 months ago, I threw together a quick & dirty (unregistered) package that I called DataMimic for generating synthetic data, based on a real dataset. It was okay for what I needed it for at the time, but it’s definitely low quality, and now I would like to revisit it and turn it into a proper package. I have had success with spec-driven development and Claude, so I want to take that approach here, but I want to leverage the incredible human expertise in this community to get feedback on the spec:

> <https://github.com/mthelm85/DataMimic.jl/blob/main/PACKAGE_SPEC.md>

I’m particularly interested in:

1. The Lux over Flux logic
2. The type system design
3. The proposed public API

And then anything else folks think should be changed/improved/etc. Thanks!!

---

<div class="post-metadata">

**Author:** ![mthelm85](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/mthelm85/32/224164_2.png) [@mthelm85](https://discourse.julialang.org/u/mthelm85)\
**Post date:** [September 16, 2026, 2:18pm UTC](https://discourse.julialang.org/t/package-technical-spec-feedback/139056/2 "2026-09-16T14:18:48Z")

</div>

I would have loved to get some feedback on this, but I went ahead and developed it anyways:

> **[GitHub - mthelm85/DataMimic.jl: DataMimic generates a synthetic DataFrame that...](https://github.com/mthelm85/DataMimic.jl/)**
>
> DataMimic generates a synthetic DataFrame that mimics the shape, column types, and statistical distributions of an input DataFrame.

If you have the need to generate synthetic data, take this for a spin and let me know what you think!