Experience

My professional journey in development and research

Full Stack AI/ML Engineer

Invene

February 2022 - Present
McKinney, TX

Leading AI/ML development for healthcare monitoring solutions and medical imaging applications.

Key Achievements:

  • Led development of real-time EEG stress/fatigue detection model deployed in ICU monitoring devices, achieving >90% precision and <200ms inference latency
  • Built deep learning segmentation pipeline for brain MRI scans using U-Net, reducing manual annotation time by radiologists
  • Engineered edge deployment system for cognitive state models on embedded Linux ICU hardware using Docker and PyTorch optimizations
  • Developed clinician-facing dashboards using Plotly Dash to visualize EEG waveform states and segmentation masks for assisted diagnosis

AI/ML Engineer

Axxess Technology Solutions

October 2020 - January 2022
Dallas, TX

Developed machine learning solutions for healthcare and financial applications.

Key Achievements:

  • Built an NLP-based email triage system using fine-tuned BERT models to auto-classify support inquiries, saving 30+ hours/week in manual review
  • Developed a model explainability layer using SHAP and LIME for loan default predictions, aiding compliance and analyst transparency
  • Deployed scalable ML pipelines on GCP with FastAPI and Docker, supporting continuous retraining to adapt to data drift
  • Collaborated with operations and support teams to integrate AI tools into production workflows with minimal latency overhead

Software Engineer

Iodine Software

June 2018 - September 2020
Austin, TX

Focused on machine learning solutions for logistics and route optimization.

Key Achievements:

  • Built an ETA prediction engine using historical delivery and GPS data, reducing ETA error from 30min to <10min through hybrid ML + ruleset approach
  • Developed LSTM-based truck demand forecast model for regional logistics, improving fleet utilization by 18%
  • Designed a geospatial clustering system for route optimization, cutting dispatch planning time and fuel consumption using HDBSCAN and OpenCV
  • Integrated ML systems into Node.js-based dashboards and automated inference pipelines using AWS Lambda

Data Science Intern

Regional Healthcare Analytics Firm

January 2015 - May 2015
Beaumont, TX

Applied machine learning to healthcare data for patient outcome prediction.

Key Achievements:

  • Built a logistic regression model to predict 30-day readmission risk using EHR data, with ROC-AUC evaluation for deployment at partner hospitals
  • Designed ICU mortality prediction models and collaborated with clinicians to validate medically relevant predictors
  • Created interactive Jupyter dashboards for hospital leadership to monitor patient risk trends and care gaps
  • Used Python, Scikit-learn, Pandas, and Jupyter Notebooks for medical code normalization and analysis

Blake Sonnier

Full Stack Developer & Machine Learning Enthusiast based in Boston, specialized in creating modern web applications and blockchain solutions.

Connect

Keyboard Shortcuts

K
Search
T
Toggle theme
2025 Blake Sonnier. All rights reserved.
Built withusing Next.js & Tailwind