User researchCAN I VIBECODE
Productlane?
BUILD FIRST
- One specific customer type and one primary use case
- A vertical slice that completes the main job without manual workarounds
- Visible limits, review steps, and a fallback when an external dependency fails
LEAVE FOR LATER
- Broad third-party integrations
- High-scale collaboration, syncing, and permissions
- Automation that can cause harm without human review
RESEARCH NOTES
What this tool actually does.
Productlane is a B2B customer support and feedback platform built on Linear, founded 2022 in Munich, Germany (~6 employees, ~$825K funding). It consolidates customer conversations from email, Slack, Teams, and live chat into one inbox, with an AI agent that resolves routine questions grounded in the company's own sources. It markets itself as the first AI support agent that interacts with Linear: it creates or links Linear issues with full context and can assign small bugs to a coding agent. A customer-facing portal combines a feedback board, roadmap, and changelog, all bidirectionally synced with Linear (projects, issues, initiatives). Users submit ideas, upvote, and comment; each idea maps to a Linear issue or project. The company positions it as connecting user research and feedback to close the loop, targeting PMs, designers, and user researchers.
RESEARCHED BUILD PROMPT
Give this to your coding agent.
Build a focused replacement for Productlane's core workflow: a support-and-feedback hub whose source of truth is an external issue tracker (Linear or GitHub Issues). Workflow: (1) Ingest customer messages from email and an embeddable web widget into one unified inbox; (2) an AI triage agent classifies each message as routine question or feedback/issue, drafts a reply grounded only in uploaded docs (no training-data knowledge), and either resolves it or creates a linked issue with full context; (3) a public portal shows a feedback board, roadmap, and changelog synced bidirectionally with the tracker, so upvotes and comments on ideas map to real issues; (4) closing an issue publishes it to the changelog. User-facing requirements: unified inbox with per-channel badges, AI-drafted replies requiring human approval before send, feedback board with upvote/comment, roadmap and changelog pages, and admin settings for AI source docs and sync mapping. Technical requirements: Node.js/TypeScript backend, React frontend, Postgres, webhooks for bidirectional tracker sync, OpenAI-compatible API for the agent with retrieval over uploaded docs, and OAuth login. Implementation approach: call the tracker's real API as the system of record and store only sync metadata locally; do not reimplement issue tracking. Do not copy Productlane's brand, logo, copy, or UI; design an original minimal interface. Out of scope: multi-tenant billing, native Slack/Teams apps (email and widget only), mobile apps, analytics dashboards, and production-grade AI safety guardrails. Include a README covering setup, env vars, sync architecture, and limitations (AI replies need review, sync conflicts, tracker rate limits).