> Markdown version of [/jobs/ext/4588489-lead-data-scientist](https://www.wearedevelopers.com/jobs/ext/4588489-lead-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Lead Data Scientist

- **Company:** Firefly Neuroscience, Inc.
- **Location:** United States (Remote available)
- **Experience:** Expert
- **Contract:** Permanent contract
- **Skills:** Amazon Web Services, Clinical Data Repository, Python (Programming Language), MATLAB, Machine Learning, NumPy, Reference Data, SciPy, Signal Processing, Verification and Validation (Software), Management of Software Versions
- **Published:** November 5, 2026
- **Apply:** https://arc.dev/remote-jobs/j/redirect/pwob07otdi

## About the Role

Required

 * 8+ years in a data science role.
 * Master's degree or higher - or equivalent sustained hands-on research orindustry experience - in neuroscience, neuroimaging, clinical neurophysiology,biomedical engineering, or a closely related field, with enough depth to reasonindependently about what a brain measure reflects physiologically and whereit breaks down.
 * Demonstrated experience building, shipping, and maintaining machine-learning or statistical models on neural or clinical data (EEG, MRI, or otherneuroimaging or physiological modalities), including sound handling of smallsamples, leakage, and class imbalance.
 * Hands-on experience building and validating normative or reference datasets:defining reference populations, covariate adjustment, z-score construction,and documenting limitations.
 * Strong production Python (NumPy/SciPy/MNE or equivalent) as your primarylanguage, with hands-on cloud experience (AWS preferred; GCP or Azureequivalent accepted). You have deployed and maintained analytic models inproduction at least twice, including monitoring, versioning, and fixing themafter release. MATLAB-only or on-prem-only backgrounds will not fit.
 * Ability to take over analysis code you did not write, re-derive the math, findwhere it is wrong, and fix it with a documented before and after.
 * Comfort operating with high autonomy, limited process, and directaccountability for outcomes.

Preferred

 * Hands-on EEG or electrophysiology signal processing: filtering, re-referencing, artifact rejection, spectral estimation, and ERP extraction.
 * Clinical research and biostatistics for external sponsors: stratified analyses,multiple-comparison correction, and methods write-ups.
 * Prior work on an FDA-cleared or otherwise regulated software product,including design controls or verification and validation documentation.
 * Experience with self-supervised or foundation-model approaches on time-series or neural data.
 * Peer-reviewed publications on neural or clinical data.

## Description

* Build, validate, deploy, and monitor machine-learning models on real clinicaldata - biomarker models, calibrated replacements for rules-based logic, andthe foundations for models trained at scale on our EEG corpus.
 * Own the measures. For every number on a Firefly report, own what it is, how itis computed, what reference it is compared against, and its documentedlimitations.
 * Verify and strengthen the existing analysis pipeline: reproduce and auditinherited methods, set acceptance criteria before comparisons, and fix whatdoes not hold up - with evidence that would stand up to clinical and regulatoryreview.
 * Build and maintain the normative reference data our reports compare patientsagainst: reference population definition, covariate adjustment (age, sex), z-score construction, sample-size adequacy, and a process for refreshingnorms as data accumulates.
 * Deliver sponsored-research analyses for external partners on time and to astandard that survives peer review, stating plainly what a sample can andcannot support.
 * Partner with the Tech Lead on delivery and with our clinical advisor on clinicaldirection; ship through the engineering team's process, including changecontrol for any change to a measure or norm on the product.