DOCUMENT INDEX
Click to unlock more detail. The more you open, the more you learn about my method and context.
Executive Brief
Andrea Zedda
PhD Bioengineer · R&D ML Architect
I design data-driven decision systems: causal inference + interpretable predictive modeling for business intelligence, and sensor-fusion motion analytics for bioengineering.
- Decision intelligence: interpretable ML + driver analysis (SHAP)
- Causal discovery/inference for decision support and strategy
- Motion analytics: sensor fusion and biomechanical metrics (IMU/EMG)
Principles & Methods
Governance boundaries, operating system, and the methods used to turn data into decision-ready outputs.
Operating System CORE
Collaboration Modes UNLOCK
Focus Now UNLOCK
What I'm building right now and the methods I use to turn data into decision-ready outputs.
Capabilities UNLOCK
Services are scoped as deliverables: reproducible analysis, decision support, and production-ready interfaces (BI / health / wearables).
Toolbox UNLOCK
Future Plans UNLOCK
What I'm building and where I'm investing my time.
Verifiable Artifacts & Context
Roles, containers, outputs, public artifacts, reading list, and academic coursework.
About CORE
A compact timeline of research and industry work, with the tooling that shaped my approach.
Operational & Research Contexts UNLOCK
Listed for context; exposure does not imply employment or endorsement.
DIRECT WORK
- SketchData
- YouAndTech
- University of Cagliari
- University of Genoa
- Camelot Biomedical Systems
OPERATIONAL EXPOSURE
- Motorola
- ITA Airways
- Aura (early-stage startup)
Education UNLOCK
Degrees and training that shaped my research approach and engineering standards.
PhD in Biomedical Engineering
University of Cagliari • 2017–2021
- Home neuromotor telerehabilitation system for stroke patients (DoMoMEA).
- Low-cost sensing + gamified feedback loops for adherence and correctness.
- Remote personalization + continuous progress monitoring.
- Android TV box + Unity3D for the patient UI.
- IMU sensing (MUSE by 221e) + custom pressure sensors.
- Multi-device Bluetooth Classic connectivity.
- Store-and-forward telemonitoring with Django + PostgreSQL.
- Nginx + Gunicorn for serving the web app.
Containers UNLOCK
A container-first ecosystem: each pillar encodes governance boundaries, reusable assets, and the current status of delivery.
Project Examples UNLOCK
Concrete builds that show what I actually ship inside the containers above.
Selected Proof
Links to public artifacts. Placeholders are intentionally marked TODO until a public URL exists.
Proof (artifacts) UNLOCK
- Android TV box + Unity3D patient-facing system for home therapy.
- IMU sensing (MUSE by 221e) + custom pressure sensors (insole/glove/under-elbow).
- Multi-device Bluetooth Classic connections to the TV box.
- Store-and-forward upload + clinician review via Django/PostgreSQL (Nginx/Gunicorn).
- Per-repetition metrics (angles, timing, errors, score) stored for longitudinal tracking.
- Clinician-configurable safety limits + remote configuration updates.
- Development status in 2019: 2/5 exercise blocks integrated; tuned on healthy subjects.
- Pilot RCT plan: 8 weeks, 5 days/week; IPPA/PIADS/QUEST/SUS for usability/acceptability.
- Engineering issue: Unity↔Android constraints; coroutine-based workaround + lifecycle mitigation.
Reading
Reading is organized by domain; each book includes why it matters (what capability it shaped).
Reading (why it matters) UNLOCK
University Courses
Exams and courses completed during my academic journey.
Courses (details) UNLOCK
Contact
Best channel: email. If you're evaluating collaboration, LinkedIn is the fastest context sync.