KNOWN ISSUES
No known issues right now - smooth sailing.
HOW IT WORKS

Smart IT ticket routing,
without the guesswork.

Pathfinder reads your issue, figures out which team should handle it, and creates the ticket — all in under 30 seconds.

24,270 Learned Patterns
7 Active Teams
4 Tickets Routed
25% High Confidence

From issue to ticket in 5 steps

No forms to fill, no teams to guess. Just describe what happened.

1

Describe the issue

Give it a title and explain what happened. Add a screenshot if it helps. That's it.

Privacy
Pathfinder checks for sensitive info (credit card numbers, passwords, national IDs) and blocks submission if found. Your personal data never gets stored.
2

Clarity check

Pathfinder reads what you wrote and decides: is this clear enough for a tech team to act on?

Clear
Skip ahead to confirmation
Unclear
A few quick questions
3

Clarifying questions

If needed, Pathfinder asks 1-2 follow-up questions with suggested answers. It uses AI to understand context, not just keywords.

AI Powered
Uses Claude to generate smart, contextual questions. Never more than 3 rounds.
4

Confirm the routing

See which team Pathfinder picked and how confident it is. You can override and choose a different team if you know better.

Override rate
25.0% of tickets get manually reassigned
5

Ticket created

A Jira ticket lands on the right board. You get a direct link to track it. Done.

Significant issues
Mark an issue as significant and it appears in the ticker bar at the top of the screen for 72 hours.

How Pathfinder picks the right team

Not AI. Not a dropdown. A pattern-matching engine trained on thousands of real Jira tickets.

1. Keyword Overrides

Checked first. Admins can say "anything with 'Salesforce' goes to the SF team" — these get a massive score boost (+10x) and take priority over everything else.

2. Pattern Matching

Each word in your issue is compared against 24,270 learned phrases. Phrases unique to one team score higher (IDF weighting). Multi-word matches get a bonus (trigrams +20%).

3. Confidence Score

Pathfinder measures the gap between the top team and the runner-up. Big gap = high confidence. Close race = low confidence. Below threshold = fallback to general queue.

4. Result

Returns the winning team, confidence level (high/medium/low), and scores for every team — so admins can see exactly why a ticket was routed where it was.

Why pattern matching instead of AI?

Pattern Matching
  • Same input always gives same output
  • Milliseconds, not seconds
  • Transparent — admins see the scores
  • Works offline, no API costs
  • Can't hallucinate or make things up
LLM / AI Routing
  • May give different answers each time
  • 1-3 seconds per request
  • Black box — hard to explain decisions
  • Requires API, costs per request
  • Can confidently route to wrong team

AI (Claude) is used only for the clarifying interview — the one place where understanding nuance actually matters.

What's under the hood

A single Flask app with three main areas: the user-facing form, the JSON API, and the admin dashboard.

DATA SOURCES
Jira Ticket creation + historical data
SQLite 24,270 patterns, 4 tickets
Claude API Vagueness detection + interview
PATHFINDER ENGINE
Router Pattern matching + IDF scoring
Interviewer AI-powered follow-up questions
PII Detector Blocks sensitive data
OUTPUTS
Jira Tickets Created on the right board
PostHog Usage analytics + funnels
Admin Dashboard Monitoring + tuning

Technology Stack

Py Python 3.11
Fl Flask 3
Sq SQLite (WAL)
Cl Claude API
Ji Jira REST v3
Ph PostHog
Gu Gunicorn
Do Docker

What Pathfinder knows

Everything lives in a local SQLite database. Here's what's inside.

Patterns

24,270 rows

Phrases learned from historical Jira tickets, scored by how strongly they predict a specific team. Updated nightly.

phrase board_key score match_count

Tickets

4 rows

Every ticket routed through Pathfinder. Tracks which team it went to, confidence, and whether it was overridden.

jira_key board_key confidence was_overridden

Boards

7 rows

The teams that receive tickets. One board is marked as the "fallback" for issues that don't match any patterns.

jira_key name is_fallback is_active

Keyword Overrides

Admin-managed

Manual rules that boost or suppress specific terms. These override the learned patterns when present.

keyword board_key boost
Patterns last updated: 2026-02-07 10:29:40

Where tickets get routed

These are the 7 teams currently receiving tickets through Pathfinder.

B2BSUPPORT
B2B Support
Operational support intake for B2B channels. Issues from B2B Agents, API Partners, Markets.
4,913 patterns
SF
CRM & Email
Salesforce CRM, email marketing, customer data platform issues
3,441 patterns
NH
Digital Products
Main delivery board for B2C/B2B web product development. Technical issues, UX improvements, new features.
6,119 patterns
PC
Polar Global Core
Core backend/platform work supporting web and partner flows. Backend dev, PG client, BDA dependencies.
3,183 patterns
SUP
Service Desk - Support
General IT support and service desk (fallback for unroutable tickets)
No patterns yet
SQSF
Squad - Salesforce
Salesforce squad board (test fallback)
No patterns yet Fallback
HWM
Web Marketing
Web content and marketing execution for B2X sites. Content issues, imagery, web publishing/merchandising.
6,614 patterns

Monitoring and tuning

The admin panel gives visibility into how Pathfinder is performing and tools to improve it.

Metrics

Tickets routed this week, submissions to Jira, drop-off funnel, PostHog analytics with date range filters.

Boards

See which teams are active, how many patterns each has, set fallback boards, activate or deactivate teams.

Model Health

Pattern distribution, confidence trends, override analysis, top patterns per board. See if routing is getting better or worse.

Routing Log

Every ticket ever routed — filter by team, confidence level, overrides. Click any row to see the full score breakdown.

Keywords

Add or remove keyword boosts and suppressions. Fine-tune routing without retraining — changes take effect immediately.

Live Tuning

Test routing on sample text, see how scores change in real-time, and add overrides to fix misroutes before they happen again.

How Pathfinder protects your data

Security measures already in place, and what's planned before full production rollout.

PII Detection & Blocking

Personal data (credit cards, phone numbers, national IDs, email addresses) is detected both in the browser and on the server. Submission is blocked if PII is found. Sensitive data is never stored.

CSRF Protection

All state-changing requests (ticket submission, feedback, admin operations) require a cryptographic token that proves the request came from the Pathfinder app itself, not a malicious third-party site.

Rate Limiting

API endpoints are rate-limited to prevent abuse. Ticket submission is capped at 5 per minute per user. Interview and routing endpoints have similar limits.

Injection Prevention

All database queries use parameterised statements (preventing SQL injection). All user-generated content is auto-escaped in templates (preventing cross-site scripting).

Non-Root Container

The Docker container runs as a restricted user with minimal permissions. Health checks are configured for monitoring by Azure App Service.

No Data Retention Risk

Pathfinder stores routing patterns and ticket metadata only. No personal data, no passwords, no customer information. The PII filter ensures nothing sensitive enters the database.

Planned for production rollout

The following will be implemented as part of the production deployment on the internal network.

Authentication & Access
  • SSO integration via Azure AD / Entra ID
  • Admin access controlled by SSO group membership
  • User identity from trusted SSO headers
Monitoring & Hardening
  • Audit logging for all admin operations
  • Content Security Policy headers
  • Secrets managed via Azure Key Vault
  • Dependency pinning and vulnerability scanning

How confident is Pathfinder?

Across all 4 tickets routed, here's the confidence distribution.

25%
50%
25%
High (>60%) — 1 tickets
Medium (30-60%) — 2 tickets
Low (<30%) — 1 tickets

Common questions

Why did it route my ticket to the wrong team?

Either the patterns haven't learned your specific language yet, or the description was missing key details. Admins see overrides in the dashboard and can add keyword boosts to prevent it happening again.

Why is it asking me questions when my issue is obvious?

Pathfinder is being cautious — it hasn't seen enough issues with that specific wording to be confident. The questions help both routing accuracy and the team receiving the ticket.

How often do the patterns update?

Every night. Pathfinder re-reads recent Jira tickets and learns from them. Keyword overrides take effect immediately.

Can admins fix bad routing?

Yes, several ways: add keyword overrides (boost or suppress specific terms), review patterns per team, test routing on sample text, and check the override analysis to spot trends.

What if the AI asks weird questions?

Claude is asked to mimic quick, specific tech support questions with suggested answers. But it can be imperfect. You can always skip the interview and go straight to confirmation.

Is my data safe?

Pathfinder blocks personal data (credit cards, passwords, national IDs) before submission. It checks on your device AND on the server. Sensitive data is never stored.

What's the ticker bar at the top?

Significant issues scroll across the top of the screen for 72 hours. It's like a news ticker for IT — everyone sees what's happening. Items auto-expire after 72 hours.