Ads That Think
Autonomous AI agents analyzing campaigns, flagging anomalies, and optimizing budgets across Meta and Google Ads—24/7, without manual intervention.
Real-time
Autonomous
Powered
What the AI Actually Does
Autonomous agents that handle the work, so your team handles the strategy.
Autonomous Analysis
Claude-powered agents continuously analyze campaign performance, budget allocation, and audience behavior across platforms.
Anomaly Detection
Instantly flag unusual patterns—budget spikes, conversion drops, audience shifts—before they become costly problems.
Budget Optimization
Agents recommend and execute budget shifts in real-time, reallocating spend to your best-performing channels and audiences.
Creative Briefs
Generate data-driven creative recommendations and briefs that align with what's actually working in your campaigns.
Unified Dashboard
Real-time client portal pulling all insights, recommendations, and performance metrics into one coherent view.
Zero Manual Intervention
Agents act autonomously within defined guardrails—no waiting for humans to approve routine optimizations.
What This Means for Your Business
Stop reacting. Start winning. Let AI handle the work while you own the strategy.
Faster optimization turnaround
Agents make decisions in seconds, not days of manual analysis
Continuous monitoring
No holiday—your budgets are always being optimized
Manual interventions
Agents handle routine optimizations within guardrails
Unified intelligence
Meta and Google Ads working together, not in silos
The Long Game
We're not just optimizing ads. We're building the infrastructure for autonomous marketing—where AI doesn't just assist, it leads. The agents learn. They reason across platforms. They predict what comes next. And they act.
We're at the frontier, and we need sharp people who can think in systems, design for scale, and make AI work in the real world where APIs fail, rate limits bite, and users expect reliability.
Sophisticated Tech Stack
Built on production-grade infrastructure designed for scale, reliability, and AI-native development.
AI & Intelligence
- Anthropic Claude API
- Multi-agent workflows
- Agentic patterns
- Structured outputs
Backend & Data
- Convex (serverless)
- Real-time DB
- Python agents
- Cron scheduling
Frontend & Hosting
- Vercel
- Vite + React
- TypeScript
- Responsive UI
Integrations
- Meta Ads API
- Google Ads API
- Shopify
- GA4
- Klaviyo
- Resend
Why This Stack?
Convex gives us real-time collaboration and instant API endpoints. Claude powers our autonomous reasoning. Vercel + Vite keeps the frontend snappy. Together, they make it easy to ship AI-native features fast, scale predictably, and handle integrations with ads platforms securely. No vendor lock-in surprises—just modern tooling built for production AI systems.
The 20% That Matters
The foundation is solid. We have autonomous agents running in production. Now we need to harden it, scale it, and ship it for real.
The Challenge
- Multi-agent workflow design and optimization across Meta and Google APIs
- API security, rate limiting, and safe token management at scale
- Error handling, edge cases, and reliability under production load
- Multi-tenant architecture for scaling to many clients
- Onboarding automation and data migration pipelines
What Success Looks Like
- Agents reliably optimize across platforms without failures
- New clients onboard in hours, not days
- System scales from 10 to 100+ clients smoothly
- API integrations are locked down—zero account suspensions
- Audit trails and compliance baked in from the start
How We Work
Weeks to make real impact on the platform
Work with the founder—short feedback loops, no bureaucracy
Ongoing opportunity as the platform scales
Who We're Looking For
Required
- Production experience connecting LLMs to live APIs (Stripe, CRM, ads platforms, payments)
- Python and serverless architecture (Convex, Supabase, Firebase, etc.)
- Familiarity with Meta Ads API and/or Google Ads API—or strong API fundamentals
Nice to Have
- Convex experience
- Ad tech or marketing automation background
- Agentic AI frameworks (CrewAI, LangGraph)
- Based in AEST overlap timezone
Ready to Shape the Future of Ad Tech?
If you've tackled production AI problems and love building systems that scale, let's talk.
Tell Us About Yourself
Share a project where you connected an LLM to a live API in production:
- Ads, payments, CRM, or any integration that mattered
- What went wrong and how you fixed it
- What you learned about production AI
Your AI Experience
Experience with multi-agent or agentic systems:
- CrewAI, LangGraph, or custom agent loops
- Coordination across multiple AI decisions
- Handling failures and edge cases
No generic proposals. Feel free to use AI to help you write—we'd expect it. But make it real. Tell us what you've actually built and what you learned.
We're looking for people, not resumes. Come talk to us.
