Joe Crowley

Experimental systems Applied AI

From physical signalsto systems people trust.

I build reliable paths from bench hardware and noisy measurements to software and AI systems people can use.

See the engineering evidenceDownload resume
Petaluma, California. Open to full-time engineering work and available immediately.
Engineering chainMeasurement to decision

I work across the interfaces where measurements are acquired, transformed, checked, and turned into decisions.

  1. 01
    Physical measurementDetector, pulse, calibration
    Raw signal
  2. 02
    FPGA and DAQTrigger, digitizer, metadata
    Timed data
  3. 03
    Analysis and MLQuality control, simulation, models
    Tested result
  4. 04
    Evidence and decisionProvenance, review, next test
    Action

Operating principleFind failures at the first interface that could have created them.

164DAQ channels installed at CERN
10 kVPulser driven by custom FPGA waveform logic
~10MRecords processed across production deployments
4+Healthcare deployments supported

Selected engineering evidence

Built, tested, and deployed.

I have built hardware, firmware, analysis pipelines, and production software. The work below shows where those pieces had to meet.

DAQ and firmware

MilliQan at CERN

Engineered and installed a 164-channel data-acquisition system. Developed Verilog trigger firmware, C++ control software, Python hardware interfaces, and functional tests for CAEN digitizers.

See the CERN system update

Fast timing and RF

ALICE, ANITA, and SLAC

Assembled and calibrated PMT electronics for the ALICE FIT T0+ detector. For ANITA and SLAC T-576, wrote FPGA waveform logic and the output bitstream used to drive a 10 kV pulser.

Read the ALICE thesis

Scientific computing

CMS detector validation

Built automated validation workflows on HTCondor and Singularity, including stage checks and failed-job recovery. Analyzed petabyte-scale LHC data with Python, C++, statistics, anomaly detection, and simulation.

View detector validation code

Production software

Avant Sciences

Led the software effort as a 1099 founding engineer and technical lead. Managed contributors and owned the core algorithms, full-stack delivery, deployment, and maintenance for identity-resolution systems processing about 10 million records across more than four deployments.

See the product context

Applied AI systems

Fat Tailed Solutions

Built AI workflows that retrieve source material, call tools, preserve provenance, evaluate outputs, and keep consequential decisions under human review.

See IPStrategy

Working range

I work across the whole signal chain.

Measurement systems

Data acquisition, oscilloscopes, PMT calibration, high voltage, digitizer testing, RF signals, FPGA and Verilog, failure isolation

Software systems

Python, C++, Linux, hardware interfaces, PostgreSQL, Docker, CI/CD, HTCondor, Singularity, parallel computing

Applied AI

PyTorch, supervised and unsupervised learning, anomaly detection, embeddings, retrieval, agent tools, evaluation, provenance

Technical leadership

Production ownership, contributor leadership, electronics and computing instruction, documentation, cross-functional collaboration

Joe Crowley outdoors in Petaluma
Based in PetalumaOpen to full-time engineering roles

Building Fat Tailed Solutions

Building a company changed how I engineer.

I started Fat Tailed Solutions to build AI systems whose evidence and outputs could hold up under expert review. That meant owning architecture, product decisions, deployment, evaluation, and the gaps between them.

I am ready to bring that range to a focused full-time engineering team. I do my best work where experiments, software, and applied AI meet, and where the result has to hold up under scrutiny.

Connect with me on LinkedIn

Experience

Physics, production, and applied AI.

2018 to 2020

Cal Poly, ALICE at CERN, ANITA and SLAC

Fast-timing detector electronics, RF calibration, FPGA waveform generation, and particle-propagation simulation.

2020 to 2023

UC Santa Barbara, CMS and MilliQan at CERN

Detector DAQ, trigger firmware, scientific software, high-performance computing, ML analysis, and technical teaching.

2021 to 2023

Avant Sciences, LLC

Founding software engineer and technical lead as a 1099 contractor, working concurrently with graduate research.

2024 to present

Fat Tailed Solutions

Founder and applied AI engineer building evidence-based workflows for technical and intellectual property work.

Start a conversation

If your team works where measurements, software, and AI meet, I'd like to talk.

Connect on LinkedInDownload resumeGitHub