← All work

AI · Life Sciences

Quality Management System for Life Sciences

Quality teams in pharma manufacturing were drowning in compliance paperwork. Audit prep, deviation tracking and CAPA workflows lived in disconnected tools — making it hard to spot risk before it became a regulatory finding.

AI-enabled Quality Management — AI · Life Sciences case study cover
Role
UX Lead
Duration
5 months
Year
2019–2020
Company
Compliance Group
Domain
Life Sciences · Pharma · Regulated SaaS

01

The problem

Quality teams in pharma manufacturing were drowning in compliance paperwork. Audit prep, deviation tracking and CAPA workflows lived in disconnected tools — making it hard to spot risk before it became a regulatory finding.

02

The approach

Create the most efficient path to quality excellence by embedding AI into the QMS — turning deviation, audit and risk workflows into a proactive risk-signal layer for FDA / EU MDR-regulated manufacturers.

Core functionalities shipped

  • Deviation tracking — identifies and documents process deviations, linked to corrective actions (CAR).
  • Audit & compliance — internal audits, documentation control, validation protocols for FDA / EU MDR.
  • Risk management — embedded risk assessment and CAPA management to address non-conformities proactively.
  • AI-assisted summarisation for audit reports and deviation summaries.

Process

Stakeholder interviews with QA leads, auditors and FDA-facing teams. Mapped the deviation → investigation → CAPA loop end-to-end before touching pixels.

Tested low-fidelity flows with internal pharma SMEs, then iterated against real audit scenarios before handing off to engineering.

Design contributions

Defined the workflow IA, ran usability sessions with QA professionals, and led a team of three designers establishing scalable design patterns the rest of the QMS suite still uses.

03

The outcomes

40%

Overall efficiency increase

45%

Reduction in test planning time

60%

Reduction in summary reporting time

  • In regulated domains, the design's job is to make risk visible early — not to hide complexity.
  • AI works best when it reduces report-writing time, not when it tries to replace human judgement on a non-conformity.
  • Earned positive customer feedback from Compliance Group's FDA-facing client: "You have hit the ball out of the park."

04

Selected screens

Information architecture for the QMS suite
Information architecture for the QMS suite
CAPA record header with status progression
CAPA record header with status progression
Approvals tab with E-Signature flow
Approvals tab with E-Signature flow
Analytics layer for quality signals
Analytics layer for quality signals

Keep reading

Other case studies

Working on a similar problem?

I take on senior contract, fractional and select full-time engagements where the brief is unclear and the stakes are real.

anjani.vc@gmail.com