{ "project": { "title": "Customer Review Analysis", "icon": { "type": "emoji", "value": "🔍" } }, "prompts": [ { "title": "Customer Review Analysis", "icon": { "type": "emoji", "value": "📊" }, "config": { "settings": {}, "provider": "openai", "model": "gpt-4o" }, "messages": [ { "role": "system", "content": [ { "modality": "text", "value": "You are an expert Product Analyst specialising in customer feedback analysis. Your task is to analyse customer reviews and extract actionable insights for product development.\n\nFor each batch of reviews, provide:\n\n1. Overall Sentiment Analysis: Categorise as Positive, Negative, or Mixed with percentage breakdown.\n2. Top 5 Recurring Pain Points: Most frequently mentioned issues with impact assessment.\n3. Feature Improvement Ideas: Specific, actionable recommendations based on customer feedback.\n4. Priority Classification: Rank issues as High, Medium, or Low priority for product roadmap.\n5. Key Quotes: 2–3 representative customer quotes that capture major themes.\n\nPresent findings in a structured format that's immediately useful for product planning and feature prioritisation decisions." } ] }, { "role": "user", "content": [ { "modality": "text", "value": "Analyse the following customer reviews for {{product_name}}:\n\n- Review data: {{review_text}}\n- Focus areas: {{specific_focus_areas}}\n- Current priorities: {{current_product_priorities}}" } ] } ], "tools": [], "variables": { "product_name": { "modality": "text", "value": "", "linkedDataset": "Customer Review - Dataset", "linkedDatasetColumn": "product_name" }, "review_text": { "modality": "text", "value": "", "linkedDataset": "Customer Review - Dataset", "linkedDatasetColumn": "review_text" }, "specific_focus_areas": { "modality": "text", "value": "", "linkedDataset": "Customer Review - Dataset", "linkedDatasetColumn": "specific_focus_areas" }, "current_product_priorities": { "modality": "text", "value": "", "linkedDataset": "Customer Review - Dataset", "linkedDatasetColumn": "current_product_priorities" } }, "playgrounds": [ { "messages": [ { "role": "assistant", "content": [ { "modality": "text", "value": "Customer Review Analysis: ShopEasy Mobile App\n\n## Overall Sentiment\nMixed sentiment — 60% negative, 40% positive. Customers appreciate product selection and deals but are frustrated with core functionality.\n\n## Top 5 Recurring Pain Points\n1. App crashes during checkout (High) — direct revenue loss, cart abandonment\n2. Confusing payment process (High) — conversion rate reduction\n3. Poor search functionality (Medium) — product discovery issues\n4. Inaccurate delivery tracking (Medium) — increased customer support load\n5. Cluttered interface design (Low) — general UX degradation\n\n## Feature Improvement Ideas\n- Immediate: Fix checkout stability and implement crash reporting\n- Short-term: Redesign payment flow with progress indicators\n- Medium-term: Improve search with better filtering and AI suggestions\n- Long-term: Add wishlist functionality and fix delivery tracking integration\n\n## Priority Classification\n- High: Checkout stability, payment flow redesign\n- Medium: Search enhancement, delivery tracking accuracy\n- Low: UI decluttering, wishlist feature\n\n## Key Quotes\n- \"Love the product selection but payment process is confusing.\"\n- \"App crashes when I try to checkout.\"\n- \"Great deals but search function is terrible.\"" } ] } ], "config": { "settings": {}, "provider": "openai", "model": "gpt-4o" }, "connectedDataset": "Customer Review - Dataset" } ], "evaluators": [ { "title": "Output Completeness", "state": "active", "weight": 1, "threshold": 0.7, "value": { "type": "llm-as-a-judge", "group": "ai-powered", "value": "Goal: The analysis must include all five required output sections.\n\nChecks:\n- Overall Sentiment Analysis with percentage breakdown is present\n- Top 5 Recurring Pain Points with impact assessment are present\n- Feature Improvement Ideas with specific recommendations are present\n- Priority Classification (High/Medium/Low) is present\n- Key Quotes section with at least two representative quotes is present\n\nScoring guidance:\n- 1: Missing 3 or more required sections.\n- 2: Missing 2 required sections.\n- 3: All sections present but some are superficial or lack detail.\n- 4: All sections present with good detail; minor gaps.\n- 5: All sections fully present, well-structured, and detailed." }, "connectedDataset": "Customer Review - Dataset", "runContinuously": true, "continuousSampleRate": 1 }, { "title": "Actionability", "state": "active", "weight": 1, "threshold": 0.7, "value": { "type": "llm-as-a-judge", "group": "ai-powered", "value": "Goal: The analysis must be specific and actionable enough for a product team to act on immediately.\n\nChecks:\n- Pain points include specific impact assessment, not just labels\n- Feature recommendations are concrete and scoped, not generic\n- Priority classifications include reasoning\n- Key quotes directly support the identified themes\n- Recommendations are scoped to the provided focus areas and current priorities\n\nScoring guidance:\n- 1: Generic output — could apply to any product.\n- 2: Some specific details but most content is generic.\n- 3: Adequately specific; team could act with some clarification needed.\n- 4: Mostly specific and actionable; minor vague areas.\n- 5: Fully actionable — every insight is specific, scoped, and team-ready." }, "connectedDataset": "Customer Review - Dataset", "runContinuously": true, "continuousSampleRate": 1 }, { "title": "Response Length", "state": "active", "weight": 1, "threshold": 0.7, "value": { "type": "response-length", "group": "performance", "value": { "value": "800", "unit": "tokens", "operator": "less" } }, "connectedDataset": "Customer Review - Dataset", "runContinuously": true, "continuousSampleRate": 1 } ] } ], "datasets": [ { "title": "Customer Review - Dataset", "icon": { "type": "emoji", "value": "💾" }, "columns": [ { "name": "product_name", "settings": { "type": "static" } }, { "name": "review_text", "settings": { "type": "static" } }, { "name": "specific_focus_areas", "settings": { "type": "static" } }, { "name": "current_product_priorities", "settings": { "type": "static" } } ], "rows": [ { "product_name": "ShopEasy - Mobile Shopping App", "review_text": "App crashes at checkout... love the selection but payment is confusing... no wishlist feature... delivery tracking is wrong... great deals but search is terrible...", "specific_focus_areas": "User experience, payment flow, search functionality", "current_product_priorities": "Improving conversion rates and reducing cart abandonment" }, { "product_name": "TaskFlow - Project Management Platform", "review_text": "Kanban board is slow to load... love the integrations but setup is complex... no bulk task editing... notifications are too noisy... reports are basic compared to competitors...", "specific_focus_areas": "Onboarding experience, notification system, reporting", "current_product_priorities": "Reducing time-to-value for new teams and improving retention at 90 days" }, { "product_name": "NestSense - Smart Home Hub", "review_text": "Device pairing is unreliable... voice commands miss about 30% of the time... love the energy dashboard... automations reset after updates... app UI is confusing for new users...", "specific_focus_areas": "Device reliability, voice recognition, automation stability", "current_product_priorities": "Reducing support tickets and improving first-week activation rate" }, { "product_name": "APIfy - Developer Integration Platform", "review_text": "Docs are outdated... rate limits hit without warning... love the webhook system... sandbox environment breaks often... error messages are not descriptive enough...", "specific_focus_areas": "Documentation quality, error handling, sandbox reliability", "current_product_priorities": "Reducing integration time for new developers and improving API uptime" } ] } ] }