WWII? Fabius did this to Hannibal more than a thousand years ago. The core of his strategy was to dunk on supplies, stall, and have the Carthaginians run out of food.
I imagine Iran, Ukraine, and Russia all know about Fabian strategy.
It returns current job data without requiring an account.
I built this because job data ingestion tends to become the same maintenance problem repeatedly:
- scraping HTML pages
- fixing parsers after frontend changes
- mapping inconsistent fields
- deduplicating the same postings across sources
Trace takes a different approach.
It ingests public machine-readable job feeds, translates them into a consistent format, and exposes the records through an API.
The core idea is: do not guess.
If a source provides a structured field, we normalize it where the mapping is clear. If a field is ambiguous or unavailable, the original value is preserved.
The goal is clean, predictable data without forcing our own interpretation layer into the pipeline.
No scraping.
No LLM extraction.
No inferred metadata.
No rewriting upstream intent.
I’d appreciate feedback from anyone interested in the problem and especially people who have built:
- job boards and aggregators
- labor market analytics tools
- search/RAG systems
- data ingestion pipelines
Especially interested in how others approach source reliability, normalization, and preserving meaning when upstream data is inconsistent.
A note of thanks: this project exists because of the support of my family and friends, the people I've learned from in IRC communities, and the broader open source community. I am grateful to everyone who builds, maintains, documents, and shares the tools and standards that make projects like this possible.
I'm just about to launch a job posting data api with postings I aggregate and very lightly normalize (https://kaleh.net/trace)
Anyone have an idea how this might impact me? They're not my postings, I just package em up and ship em. Strive to comply with all laws and TOS and not trying to make trouble.
So this could be used as a streamlined way for client devices (mobile phones for example) to phone home to servers (google.com for example) with user data and bypass some local network controls? (DNS block lists, for example)
looks like I installed a package 3 weeks before it was comprised. I try to limit surface area as much as possible, and this package seemed to be required for a project I'm working on. (turns out it wasn't, but i didn't know that yet)
Lol, I forgot most ppl have electronic, motorized, smart mirrors. Here I am bringing my manual mirror bias to the table thinking that ppl set their mirrors by hand.