Customer-Reported Issues → Persona Pain Points

Pull customer-reported issues from Jira via JQL, analyze pain points with Mave Agent, and create pain-point personas

Customer-Reported Issues → Persona Pain Points

Scenario

Your support and engineering teams label customer-reported issues in Jira, but that pain signal never reaches marketing. This job uses JQL to find issues labeled “customer-reported”, extracts summaries and descriptions, sends the aggregate to Mave Agent for pain-point analysis, then creates enriched personas representing common customer pain patterns.

Flow: Jira POST /search (JQL) → aggregate issues → Mavera POST /api/v1/mave/chat (pain point analysis) → POST /api/v1/personas (pain-point personas)

Code

Python

import os, requests, time, base64

DOMAIN, EMAIL = os.environ["JIRA_DOMAIN"], os.environ["JIRA_EMAIL"]
TOKEN, MV = os.environ["JIRA_API_TOKEN"], os.environ["MAVERA_API_KEY"]
JB, MB = f"https://{DOMAIN}.atlassian.net/rest/api/3", "https://app.mavera.io/api/v1"
cred = base64.b64encode(f"{EMAIL}:{TOKEN}".encode()).decode()
JH = {"Authorization": f"Basic {cred}", "Content-Type": "application/json"}
MH = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"}

jql = 'labels = "customer-reported" ORDER BY created DESC'
issues, start = [], 0
while True:
    r = requests.post(f"{JB}/search", headers=JH, json={
        "jql": jql, "startAt": start, "maxResults": 100,
        "fields": ["summary", "description", "issuetype", "priority", "status"],
    })
    if r.status_code == 429:
        time.sleep(int(r.headers.get("Retry-After", 30)))
        continue
    r.raise_for_status()
    data = r.json()
    issues.extend(data.get("issues", []))
    if start + data["maxResults"] >= data["total"]:
        break
    start += data["maxResults"]
    time.sleep(0.5)

print(f"Fetched {len(issues)} customer-reported issues")

issue_lines = []
for iss in issues[:60]:
    f = iss["fields"]
    desc = ""
    if f.get("description") and f["description"].get("content"):
        for block in f["description"]["content"]:
            for item in block.get("content", []):
                if item.get("text"):
                    desc += item["text"] + " "
    issue_lines.append(f"- [{iss['key']}] {f['summary']} | {f['issuetype']['name']} | "
                       f"{f['priority']['name']}\n  {desc[:250]}")

analysis = requests.post(f"{MB}/mave/chat", headers=MH, json={
    "message": (
        f"Pain-point analyst. Analyze {len(issues)} customer-reported Jira issues.\n\n"
        f"ISSUES:\n" + "\n".join(issue_lines)[:8000] + "\n\n"
        "Identify: 1) Top 5 pain clusters 2) Severity ranking 3) Persona archetypes "
        "4) Language patterns 5) Persona definitions (name, role, pain points)"
    ),
}).json()

PERSONAS = [
    {"name": "Frustrated Power User", "pain": "Hits edge cases and performance limits daily"},
    {"name": "Blocked Team Lead", "pain": "Integration failures prevent team workflows"},
    {"name": "Confused New User", "pain": "Onboarding gaps and unclear error messages"},
]
for p in PERSONAS:
    time.sleep(0.3)
    r = requests.post(f"{MB}/personas", headers=MH, json={
        "name": f"Pain Point: {p['name']}",
        "description": f"From {len(issues)} Jira issues. Core pain: {p['pain']}. "
                       f"Context:\n{analysis.get('content', '')[:800]}",
    })
    r.raise_for_status()
    print(f"  {r.json()['id']}: Pain Point: {p['name']}")

Example Output

Fetched 83 customer-reported issues
  per_4xM2a: Pain Point: Frustrated Power User
  per_7kN9b: Pain Point: Blocked Team Lead
  per_2jQ5c: Pain Point: Confused New User

Top 5 Pain Clusters:
1. Data Export Failures (18 issues) — Exports hang on large datasets
2. Webhook Delivery Gaps (14 issues) — Webhooks drop during deploys
3. Onboarding Confusion (12 issues) — Unclear errors, hidden settings
4. Search Performance (9 issues) — JQL queries timeout on large projects
5. Permission Errors (7 issues) — "Access Denied" on visible issues

Error Handling

JQL syntax errors (400)

Jira returns 400 with errorMessages if JQL is malformed. Common mistakes: unquoted strings with spaces, invalid field names, missing escape for reserved characters. Test JQL in Jira’s issue navigator first.

ADF description parsing

Jira Cloud uses Atlassian Document Format (ADF) for descriptions — nested JSON, not plain text. The code walks content[].content[].text. Rich content (tables, code blocks) needs deeper traversal.

Rate limit (429)

Jira uses a points-based system — search requests cost more than single-issue reads. The code reads Retry-After and waits. Add 500ms between pages for large result sets.