Generate a professional business email that communicates the message clearly and confidently. The tone should feel human, thoughtful, and well-structured—never robotic or overly formal.
What happened?
RECIPIENT
SUBJECT
CONTEXT
TONE
Generate Draft
Persuasive
Team Leads...
Project Scope...
Concise
Friendly
Professional
User Research
We spoke with business users who write emails daily.
Insight
Users don’t want better writing—they want faster judgment with control
Key findings
Users struggle to translate intent into prompts
Tone mismatches cause excessive editing
Long AI outputs reduce confidence and slow sending
User Snapshot
Busy professionals writing high-stakes emails under time pressure who value clarity, tone accuracy, and confidence before sending.
Product Bet
If intent and tone are defined before generation, AI output becomes clearer, shorter, and easier to trust.
Outcome
Users spend less time editing, feel more confident sending emails, and treat Cloud as a thinking aid.
Solution
Cloud was designed as a structured writing system, not a chat interface.
The mission
Pipeline Risk Analyst is an AI-powered agent that monitors GitLab CI/CD pipelines and merge requests in real time, automatically performs a multi-dimensional analysis using Claude (Anthropic), and posts a structured, actionable report as an MR comment — all without requiring human intervention.
Problem
Modern engineering teams merge dozens of branches per day. When a pipeline fails or a PR is opened at off-hours, developers face three compounding problems:
AI-powered CI/CD intelligence agent — powered by Claude, deployed on Google Cloud Run, native to GitLab.
Timeline
Product Design
Brand Identity
Team
2 Product Designers
Role
Lead Designer
Skills
Product Design
User Research
Prototyping
"A junior developer opens a PR at 11pm. By morning, they have a full code review, root cause analysis on the failed pipeline, and a concrete suggested fix — without waking up a senior engineer."
Delayed feedback
Context switching
Incomplete reviews