Cycling through all six. Tap any point to stop.

Measured on six things

AI-powered spatial hierarchy for Jira backlogs, read-only...

Growth

A real, specific pain (inherited 3,000-issue backlogs) with a concrete wedge: AI-inferred hierarchy, zero config, versus Structure's manual setup and Rich Filters' JQL dependency. Traction signals are thin — a LinkedIn profile and nothing else.

Screenshots

1 / 1

Teppe for Jira

About this app

Teppe for Jira is a refreshing take on an old problem: Jira backlogs that have grown into shapeless, unmanageable lists. Instead of yet another flat view or complex query, Teppe uses a combination of deterministic rules and AI clustering to automatically organise your issues into a navigable, spatial hierarchy. It's like turning a 3,000-issue backlog into a map you can actually explore. The app is read-only, so it never touches your data, and it runs entirely on Atlassian's infrastructure, meaning no external APIs or data egress. If you're a team lead, product manager, or anyone who's inherited a messy project, Teppe offers a way to finally see the forest for the trees. It's not a portfolio rollup or a reporting tool; it's a navigator for understanding what's really in your backlog.

Key features

1Zero-config spatial hierarchy: deterministic rules plus AI clustering infer a navigable structure from your existing Jira issues, no schema or admin setup required.
2Read-only, zero-egress architecture: single read:jira-work scope, runs on Atlassian-hosted LLMs, stores structural fields only — never descriptions, comments, or attachments.
3Recursive grid view: renders thousands of issues as one picture with groups capped at eight items, so you can drill in without losing context or hitting a wall.

Target customer profile

Drawn from the product itself, not from a survey.

Demographic

Team Leads & Product Managers

Pain points

Overwhelmed by large, inherited backlogs; difficulty understanding project scope; time wasted on manual triage.

Primary needs

Quick, intuitive overview of backlog structure; ability to drill down without losing context; minimal setup.

Demographic

Teams Inheriting a Project

Pain points

Backlog with thousands of issues, no clear structure, orphaned tasks; lack of documentation or context.

Primary needs

Instant visibility into work themes; identification of orphaned issues; easy onboarding to unfamiliar projects.

Demographic

Agile Teams in Planning & Grooming

Pain points

Flat lists make it hard to prioritise; epics become dumping grounds; planning sessions bogged down by navigation.

Primary needs

Thematic grouping of issues; visual hierarchy for better decision-making; efficient backlog refinement.

Scout Review

Written by AI from measured evidence, scored out of 100.

Teppe for Jira attacks a problem every Jira admin knows intimately: the backlog that grew into a shapeless list nobody can navigate. Its answer — deterministic rules plus AI clustering to infer a spatial hierarchy with zero configuration — is a genuinely different wedge from Structure (manual setup, reporting-oriented) and Rich Filters (JQL-dependent). The marketing site is fast (Lighthouse mobile 88, desktop 99, CLS 0), secure by header standards (11/12 checklist points), and accessible by automated audit (100/100). The architecture claims are specific and credible: read-only, single scope, zero egress, structural data only. The gaps are equally clear: no product screenshots anywhere on a product whose value is visual, no docs or demo, no published compliance certification, and no accessibility statement. It's a focused, well-positioned tool with a real niche — the question is whether 'install and it just works' is enough to earn trust from the exact person who has the least patience to gamble.

Measure by measure

Usability

Zero-config onboarding and a single read-only scope are genuinely low-friction, and the site promises value 'within minutes.' But there's no docs, no demo video, no changelog — just two buttons and vibes.

Design

Clean, disciplined marketing site with strong typography and a clear problem-to-solution narrative. But it never shows the actual product UI — only abstract glyphs — so the app's own design is unjudgeable from here.

Speed

Lighthouse mobile 88/100 (LCP 2.7s, CLS 0, TBT 0ms) and desktop 99/100. A genuinely light, well-built static page with no layout shift and no main-thread blocking.

Security

Strong header hygiene (11/12: HSTS, CSP, frame protection, SPF, DMARC) and a credible read-only, zero-egress, single-scope architecture on Atlassian infra. No SOC 2, ISO, or pen-test claim published.

Accessibility

Lighthouse a11y audit scored a perfect 100/100, and the site is a clean static page with readable contrast. No WCAG statement, no a11y page, and no multi-language support mentioned.

Growth

Improvement areas

No product screenshots or demo video anywhere on the site — a visual product with an invisible UI.
No docs, help center, or changelog; onboarding relies entirely on 'install it and it works.'
No published security certification (SOC 2, ISO, pen-test) and no accessibility statement or WCAG claim.

Design & Usability

Teppe's site is a well-crafted piece of marketing: disciplined typography, a clear Problem → What it does → How it works → Who it's for flow, and copy that names the pain precisely ('an epic with 200 children is common — and conveys nothing'). The problem is that for a product whose entire value is visual — a spatial hierarchy you can navigate — the site shows zero product screenshots. Four abstract glyphs (◱ ▦ ✦ ⛉) stand in for the actual grid. On usability, the pitch is genuinely low-friction: no admin, no schema, a single read:jira-work scope, and value 'within minutes.' That's a real onboarding advantage over Structure, which requires configuration. But there's no docs, no help center, no changelog, no demo video — just two buttons. For a tool aimed at people who inherited a mess and need to trust it fast, the absence of any walkthrough is a gap.

Speed & Security

On speed, Teppe's marketing site is genuinely well-built: Lighthouse mobile 88/100 with LCP 2.7s, CLS 0, and TBT 0ms, and desktop 99/100. A CLS of zero and zero main-thread blocking means a lightweight static page with no layout thrash — exactly what you want from a landing page. On security, the HTTP probe returns 11/12 checklist points: HTTPS enforced, HSTS, CSP, frame protection, X-Content-Type-Options, Referrer-Policy, plus SPF and DMARC. Only Permissions-Policy is missing. The architecture claims are credible and specific: read-only, a single read:jira-work scope, zero egress, AI clustering on Atlassian-hosted LLMs, and structural fields stored but never descriptions, comments, or attachments. That's a thoughtful data-handling posture. What's absent is third-party validation — no SOC 2, ISO 27001, or pen-test report is published, which is common for an early-stage Marketplace app but leaves the security story resting on the maker's own word.

Accessibility & Growth

Accessibility is the strongest measured dimension: Lighthouse's axe-core audit scored 100/100, and the site is a clean static page with readable contrast and no obvious keyboard traps. But automated checks cover only part of WCAG, and Teppe publishes no accessibility statement, no WCAG conformance claim, and no multi-language support — so the ceiling is set by what wasn't documented, not by a flaw. On growth, the idea holds up. The problem is real and specific: a 3,000-issue backlog has knowable size but no knowable structure, epics become dumping grounds, and orphaned issues are invisible to every structural view. The target customer is named precisely — team leads and PMs who inherited a backlog they didn't create, and teams post-reorganisation or acquisition. The wedge against the listed alternatives is concrete: Structure for Jira requires configuration and is built for reporting, Rich Filters relies on JQL and doesn't infer structure, and Jira's native hierarchy needs manual maintenance. Teppe's zero-config, AI-inferred hierarchy is a genuine differentiator. The limit is that this is a single-feature app inside Atlassian's ecosystem, with no network effects and no data moat — Atlassian could absorb the capability, and the addressable market is bounded by Jira Cloud teams with large messy backlogs. Traction signals are thin: a LinkedIn profile and nothing else in the collected evidence.

Conclusion

Teppe is the rare Jira app that doesn't try to be a portfolio rollup or a reporting suite — it picks one job, backlog comprehension, and does it with a defensible technical approach. The read-only, zero-egress, single-scope design is the right call for a tool that touches sensitive project data, and the measured performance and accessibility numbers back up the craft. What would move this from promising to proven: a 60-second demo video showing the actual grid on a real 3,000-issue project, a docs page, and a published security posture (even a self-attestation or pen-test summary). For a team lead staring down an inherited backlog, Teppe is worth a Marketplace trial — just go in knowing you're trusting the pitch, because the receipts aren't public yet.

Suggestions for improvement

Add a 60-second demo video or annotated screenshots showing the actual recursive grid on a real large backlog — the product's value is visual and the site shows none of it.
Publish a docs/help center page covering setup, the clustering logic, and what 'structural data only' means in practice, to reduce trust friction at install.
Add an accessibility statement and WCAG conformance level, and consider multi-language support for international Jira teams.
Publish a security posture page — even a self-attestation, pen-test summary, or roadmap toward SOC 2 — to back the read-only, zero-egress claims with something verifiable.

How we score

Alternatives comparison

Named competitors, point by point. Nobody paid to appear here or to be left out.

	Teppe for Jira	Jira's Built-in Hierarchy	Structure for Jira	Rich Filters for Jira
Setup effort	Zero config — install and it builds the hierarchy automatically	Manual — requires admins to build and maintain epic links	High — requires configuration and hierarchy schema setup	Medium — requires writing JQL and building dashboards
Automatic structure inference	Yes — AI clustering groups issues by meaning, not just labels	No — structure exists only where a human created it	No — you define the hierarchy manually	No — views are driven by JQL queries you write
Primary use case	Exploration and comprehension of large inherited backlogs	Core issue tracking and native epic hierarchy	Portfolio rollup and reporting across projects	Dynamic filtering and dashboard reporting
Data access model	Read-only, single read:jira-work scope, zero egress	Native — full read/write within Jira itself	Read/write — stores hierarchy data in its own layer	Read-focused — renders filtered views of Jira data
Learning curve	Low — no JQL, no schema, value within minutes	Moderate — familiar to Jira users, manual structure upkeep	High — powerful but configuration-heavy	Moderate to high — requires JQL proficiency