Building a Leadership Feedback & Coaching Tool at Work.Life

What they built

A custom web-based Leadership 360 feedback and coaching tool that allows all Work.Life employees to rate and review each member of the Senior Leadership Team (SLT) across seven leadership parameters, with AI-generated coaching reports grounded in real frameworks.

Key features:

• Public-facing URL with email-based vetting (checks against HR platform to confirm active employees only)
• Optional anonymity for reviewers — name is not required, but email is used for verification
• Seven quantitative rating parameters plus open-text feedback fields
• Admin panel to manage leaders, toggle participation per round, and create/close feedback rounds
• Round-based structure with unique invite links per round
• Live response tracking (completions, reviews submitted per round)
• Auto-generated coaching reports per leader, editable before sharing

Why they built it

• Quarterly engagement surveys showed leadership scores were stubbornly low while all other categories improved
• Standard engagement surveys lacked the depth to diagnose why — a 360 format could surface more specific, actionable insight
• Wanted to give employees a structured, psychologically safe way to give upward feedback
• Needed coaching-quality output that could be shared directly with SLT members' external coaches or used in performance and growth conversations

Tools used

• Claude - wrote the main bulk of the application code
• Lovable - used to build the final UI layer ("the last mile") from the code Claude produced
• HR platform integration - email vetting cross-references the company's existing HR system to validate reviewers
• Coaching frameworks embedded in the tool:
  • Perception Gap and Blind Spot Analysis Framework
  • Strengths-Based Executive Debrief Framework

How you could build this yourself

1. Define your leadership parameters - identify the 5–7 dimensions of leadership you want to assess (e.g. communication, vision, psychological safety). Align these to a recognised coaching framework so questions feed directly into the analysis.
2. Choose your coaching framework(s) - select real, named frameworks (e.g. Strengths-Based Executive Debrief, Perception Gap Analysis). Document how each survey question maps to each framework dimension.
3. Draft your prompt logic in Claude - write detailed instructions explaining how the framework works, how questions connect to it, and strict guardrails (e.g. only use submitted data; do not infer or hallucinate context). Ask Claude to generate the application code.
4. Build the UI in Lovable - paste or import the Claude-generated code into Lovable to create the front-end interface, including the reviewer landing page, rating screens, and admin panel.
5. Set up email-based vetting - connect the tool to your HR platform or an employee data export so only verified employees can submit reviews.
6. Configure admin controls - build in the ability to add/remove leaders, open and close rounds, and generate unique invite links per round.
7. Test anonymity settings - confirm that anonymous submissions cannot be traced back to individuals, and that SLT members cannot see their own rating card.
8. Pilot with one round - launch with a clear, supportive message framing it as a growth tool ("help your leaders improve"), not a performance judgment.
9. Review and edit AI coaching output before sharing - treat generated reports as a first draft; edit before sending to leaders or their coaches.
10. Connect outputs to existing processes - route coaching reports to external coaches, mentors, or into your performance and growth conversation cycle.

Source

Built by Olya Yakzhina, People lead at Work.Life, and presented as part of an Open Org "Shipped by HR" session. Olya is a qualified coach and has previously implemented SLT 360 processes at other organisations, including Switchee