Nelli

Web application

Nelli is a certified web-based Class IIa medical device with FDA clearance for seizure detection and characterization, enabling video and audio recordings of events to support faster diagnosis, treatment decisions, and clinical follow-up.

Role

Lead UX/UI Designer

Responsibility

Ownership of UX

Regulatory & Safety Impact
  • Supported MDR compliance and FDA 510(k) clearance by leading the UX and Human Factors process and usability documentation for submissions and audits.

  • Built a structured usability risk and traceability framework by applying IEC 62366 and ISO 14971 and documenting 300+ hazard-related use risks to guide design decisions.

  • Validated regulatory usability requirements and translated safety-critical findings into design controls through formative and summative testing in clinical environments.

Product & Workflow Impact
  • Improved clinical event review workflows through iterative, evidence-based UI design in close collaboration with engineering, clinical, and regulatory stakeholders.

  • Ensured clinically accurate and technically feasible UI solutions through continuous collaboration with engineering, data science, QA, RA, and clinicians.

  • Contributed user and risk perspectives to feature prioritization discussions with the development team.

Research & Alignment Foundations
  • Established a shared understanding of clinical user needs by developing personas and workflows during foundational user research.

  • Bridged research, development, and end-user adoption by writing user stories and epics, facilitating feedback sessions, and supporting user training.

Organizational Contribution
  • Standardized Human Factors practices across the organization by defining and implementing the Human Factors Engineering SOP and work instructions.

  • Improved coordination by leading a small development team.

Case Study 1 – Interactive Report

Core feature of Nelli providing visibility into recording periods, total seizure events, individualized classification types, and direct access to event videos.

Challenge

Enable doctors to access classified seizure events quickly and reliably, balancing ease of use with data security and clinical responsibility.

Use cases
  • Visualized event distribution

  • Individualized seizure classification

  • Aggregated totals by event and type

  • Direct access to individual events

Approach
  • Defined clinical personas

  • Engaged end users iteratively

  • Tested multiple design concepts

  • Validated workflow with usability testing

  • Added functionality incrementally

  • Performed ongoing usability risk analysis

Impact
  • 12-color classification system: enabled clinicians to instantly recognize event types for 9 epileptic and 3 non-epileptic categories, regardless of prior patient context. This system is now part of the design library.

  • Progressive disclosure for long recordings: only the current month expands while others collapse, reducing load time while maintaining access to relevant information.

  • Direct selection for multiple patient orders: allowed clinicians to jump directly to the desired recording period, improving navigation efficiency. This pattern is now part of the design library.

Case Study 2 – Continuous Monitoring

Experimental feature for continuous patient monitoring, developed in collaboration with Danish Epilepsy Hospital.

Challenge

Design a hospital-specific solution for near real-time seizure monitoring with visual and audio feedback, enabling on-call nurses to respond quickly while monitoring multiple patients simultaneously.

Use cases
  • Monitor multiple patients simultaneously

  • Display prioritized seizure events

  • Ensure patient safety

  • Protect patient privacy and data

  • Detect patient leaving bed area

Approach
  • Created dedicated persona

  • Collaborated closely with client

  • Defined use cases, epics, and user stories

  • Designed Timeline layout

  • Iterated designs based on usability testing

  • Defined distinct sound notifications

  • Performed ongoing usability risk analysis

Impact
  • Three-tier color system: red/orange/blue aligned to seizure priority, enabling nurses to immediately identify critical events and respond without missing seizures. Now part of the design library.

  • Distinct sound notifications: allowed on-call nurses to stay aware of events without constant visual monitoring, supporting patient safety during high workload periods.

  • Physical privacy switch: ensured recording paused during sensitive care moments, maintaining patient dignity and data protection.

Let's connect

© 2026 Tatiana Anagnostaki. All rights reserved.

© 2026 Tatiana Anagnostaki.

All rights reserved.