Jupyter notebook: is this job title plausible AT THIS COMPANY? (L2 — legal form, size, cardinality; PLAUSIBLE is confidence-capped without a registry)
  • Python 79.2%
  • Jupyter Notebook 20.4%
  • Makefile 0.4%
Find a file
brahm 9d1430303b Add a Jupyter notebook for title-at-this-company (L2).
Asks whether a title could exist at a given company — legal form, country
reading, size, file cardinality, industry — with a hard cap on PLAUSIBLE
confidence because no registry is queried.
2026-08-19 23:35:29 +00:00
data Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
docs Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
notebooks Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
results Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
scripts Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
src/tac Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
tests Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
.gitignore Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
LICENSE Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
Makefile Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
README.md Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
requirements.txt Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00
START-HERE.md Add a Jupyter notebook for title-at-this-company (L2). 2026-08-19 23:35:29 +00:00

Title at this company

Public: https://git.binary.ovh/brahm/title-at-company

One Jupyter notebook for a question the L1 title-string workbench cannot answer:

Given this title and this company, how sure are we the title is valid here?

Not “is this a job title?” (L1). Not “does this person hold it now?” (L3). L2: could this title exist at this company?

The honest confidence rule

IMPLAUSIBLE can be high-confidence. PLAUSIBLE cannot.

Vorstand at a GmbH is wrong in company law. Three CEOs in one file means at least two rows are wrong. CEO at a large US corp, with no officer list, is only plausible — confidence is capped at 72% because this repo never asks a registry.

5 minutes

git clone https://git.binary.ovh/brahm/title-at-company.git
cd title-at-company
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
export PYTHONPATH=src
python tests/test_tac.py
python scripts/smoke.py
python -m tac.cli "Vorstand" --company "Müller Bäckerei GmbH" --country DE --employees 8 --industry hospitality
jupyter lab notebooks/title-at-this-company.ipynb

Mac: docs/MAC.md. Placeholder list: docs/PLACEHOLDERS.md.

The only notebook is notebooks/title-at-this-company.ipynb: approaches with printed output, then the live workbench (tempo / loop / highlighted row / CSV edits / corrections), then # >>> YOUR CODE HERE.

Approaches implemented

# Approach Hard fail?
1 L1 gate (not a title → stop) yes (question N/A)
2 Legal form / statutory office yes, if form known
3 Country / industry reading of the same string no (changes rank, not always validity)
4 Size vs seniority yes (EVP at 6 people)
5 File cardinality (duplicate CEO/Chair) yes, needs siblings
6 C-suite density vs headcount yes, needs a file
7 Industry vs distinctive occupation head no (weak)
8 as-of before incorporation yes
9 Registry officer list (L3) placeholder only

Related L1 practice bench: job-title-workbench. Related production-shaped pipeline: contact-trust.