AI-native SEO agent

Your SEO team,
down to one agent.

rankr researches keywords and competitor gaps, drafts the article, and remembers what's shipped. You stay in the loop and approve every step.

Watch the demo →See it live →
✓ Human-approved·✓ Per-client memory·✓ Reddit-grounded
app.rankr / agent
09:14researching latex mattress sg competitors
09:14found 3 ranking gaps across 2 competitor domains
09:15drafting “best latex mattress in Singapore”
09:15outline · writing · CTA · review · packaged
09:16● awaiting your approval
Competitor gap
NEW latex mattress sg
Draft ready
approve to publish →
UCWS Singapore 2026
Top 20 · 800+ builders
Backed by
SUTD Baby Shark Fund
Live testbed
lullaflex.com/blog
The edge
Reddit-grounded research
01 / The problem

SEO work is scattered across five tools and three docs.

Keyword tools, competitor reports, spreadsheets, CMS drafts, chat prompts. Even with AI writing, the hard questions still sit unanswered between the tabs.

  • 01What should we write next?
  • 02Why is this topic worth prioritizing?
  • 03Which competitor gap or search demand supports it?
  • 04Have we already covered this topic?
  • 05How do we move from research to draft without losing context?
02 / How it works

One loop, run end to end. You stay at the gate.

Every client runs the same loop. The agent proposes, you approve, and whatever ships feeds the next recommendation.

01
Research
keywords + competitor gaps
02
Recommend
ranked content opportunities
03
Approve
you stay in the loop
04
Write
structured article workflow
05
Publish
draft → live
06
Remember
coverage memory
↻ coverage memory feeds the next research pass. the loop compounds per client.

Powered by 11 specialist agents, from search-demand analysis and competitor research to long-form writing and editorial review.

03 / What it does

A command center for content operations.

01
Per-client workspace
Each client has its own context, audience, brand voice, language, and location.
02
Keyword research
Finds keyword opportunities with DataForSEO and stores them as scored content opportunities.
03
Competitor intelligence
Analyzes competitors, ranking gaps, and domain intersections into gap recommendations.
04
Article workflow
Turns an opportunity into a structured draft: research, outline, writing, CTA placement, review, packaging.
05
Human approval
The agent reads context, recommends the next action, and only runs it after you confirm.
06
Coverage memory
Tracks approved and published content so future recommendations don't repeat what's covered.
built onNext.js · Supabase · Trigger.dev · DataForSEO · Gemini / LLM agents
04 / See it run

The agent, working a real client.

Generated and approved articles publish to a live blog at lullaflex.com/blog. The system is tested against a real publishing workflow, not a sandbox.

rankr demo · youtube
05 / Who built it

Two builders, one agent.

Abel Lee
AI Engineer
Abhishek Vulla
Full Stack
Built at SUTD · Singapore · backed by the Baby Shark Fund