engineeringdata

Prompt Engineering Guide

Design, test, and optimize LLM prompts systematically — with evaluation frameworks, chain-of-thought patterns, output formatting, and iteration methodology for reliable AI outputs.

promptsLLMevaluationchain-of-thoughtAIoptimization

Works well with agents

AI Engineer AgentML Engineer AgentPrompt Engineer Agent

Works well with skills

Experiment DesignTest Plan Writing
prompt-engineering-guide/
    • customer-support-classifier.md3.3 KB
  • SKILL.md6.4 KB
prompt-engineering-guide/examples/customer-support-classifier.md
customer-support-classifier.md
Markdown
1# Prompt Engineering — Support Ticket Classifier, Helios SaaS
2 
3## Task Definition
4 
5**Goal:** Classify incoming support tickets into one of 6 categories for automatic queue routing.
6**Model:** Claude 3.5 Sonnet. **Volume:** ~800 tickets/day, budget $0.01/ticket max.
7 
8### System Prompt (v4 — final)
9 
10```
11You are a support ticket classifier for Helios, a project management SaaS product.
12Given a support ticket (subject and body), classify it into exactly one category.
13 
14Categories:
15- billing: Payments, invoices, subscriptions, refunds, pricing
16- bug: Something broken, errors, crashes, unexpected behavior
17- feature_request: Requests for new functionality
18- account: Login, password resets, SSO, permissions
19- integration: API questions, webhooks, third-party connections
20- how_to: Usage questions, workflow guidance
21 
22Rules:
23- If a ticket spans two categories, choose the one requiring action first.
24- Broken integrations are "bug" not "integration."
25- If uncertain, respond with "unknown".
26 
27Respond with JSON only:
28{"category": "string", "confidence": "high | medium | low", "reasoning": "1 sentence"}
29```
30 
31## Few-Shot Examples
32 
33```
34Ticket: "Can't export CSV anymore" / "Export button throws 500 error since yesterday."
35Output: {"category": "bug", "confidence": "high", "reasoning": "Specific error on previously working feature."}
36 
37Ticket: "Add Gantt chart view" / "Would love a timeline view for projects."
38Output: {"category": "feature_request", "confidence": "high", "reasoning": "Requesting new functionality."}
39 
40Ticket: "Slack notifications stopped" / "Integration was fine until last week, nothing changed."
41Output: {"category": "bug", "confidence": "high", "reasoning": "Previously working integration is broken."}
42 
43Ticket: "How do I set up recurring tasks?" / "I want tasks to recreate every Monday."
44Output: {"category": "how_to", "confidence": "high", "reasoning": "Asking how to use an existing feature."}
45```
46 
47## Evaluation Framework
48 
49Test set: 50 manually labeled tickets, stratified by category.
50 
51| Metric | Method | Threshold |
52|--------|--------|-----------|
53| Format compliance | JSON schema validation | 100% |
54| Accuracy | Exact match vs. human labels | > 90% |
55| High-confidence accuracy | Accuracy on "high" confidence only | > 95% |
56| Latency (p95) | API response time | < 2s |
57| Cost per ticket | Token count * price | < $0.01 |
58 
59## Iteration Log
60 
61| Version | Change | Accuracy | Key Issue |
62|---------|--------|----------|-----------|
63| v1 | Category list only, no examples | 74% | Confused integration/bug (6 errors), inconsistent JSON (3 errors) |
64| v2 | Added 3 few-shot examples | 84% | how_to/feature_request confusion on "Can I do X?" tickets |
65| v3 | Added how_to example + broken-integration rule | 90% | 2 billing/account edge cases, 2 ambiguous bug/feature |
66| v4 | Added Slack edge case + confidence field | 92% | 4 remaining errors are genuine ambiguities (60% human agreement) |
67 
68**v4 accepted.** High-confidence accuracy: 96%. Format compliance: 100%. Avg cost: $0.003/ticket. Avg latency: 680ms.
69 
70## Production Configuration
71 
72- **Temperature:** 0 (deterministic)
73- **Max tokens:** 150
74- **Fallback:** "low" confidence tickets route to human triage
75- **Monitoring:** Alert if any category exceeds 2x its 30-day average volume
76- **Re-evaluation:** Monthly with 10 fresh test tickets added per cycle
77 
AgentsSkillsCompaniesJobsForumBlogFAQAbout

©2026 ai-directory.company

·Privacy·Terms·Cookies·