Piyush Barik

All case studies

Lancaster University · 2025

SAIL: Safe AI at Lancaster

Turning a high-trust university AI tool into something people could use without second-guessing it.

Visit the live project

SAIL: Safe AI at Lancaster interface
SAIL: Safe AI at Lancaster · public case study, ai, platform

Role

Full-stack Developer

Year

2025

Stack

  • React
  • Vite
  • TypeScript
  • .NET Aspire
  • Azure AI Foundry
  • REST APIs

Overview

SAIL gives the university community access to modern AI tools inside a secure, policy-aware environment. It had to feel capable without looking experimental or risky.

Challenge

  1. Make AI features approachable for a broad academic audience.
  2. Reduce friction around trust, safety, and document handling.
  3. Keep the interface clear while the system underneath did sensitive work.

Approach

  1. Focused on interface clarity, progressive disclosure, and readable interaction states.
  2. Helped shape flows around document-aware tasks and knowledge-driven interactions.
  3. Worked across the stack, on the interface and supporting services, fitting into a larger platform.

Outcomes

  1. Delivered a clearer experience for a sensitive, fast-moving product area.
  2. Met the institution’s trust requirements without making the interface feel locked-down.
  3. Left something that felt credible to staff, students, and researchers.

Curious about the longer version, or the parts that didn't make the page? I'd be happy to talk it through. Get in touch