5 REAL AI Agents We’ve Built in the Last 6 Months

---

---

AI Summary

Video Summary: Real Production AI Agents in Action

Agent Use Cases

1. Facebook Marketing Agent for Carab
   • Carab is a Dutch online learning platform for nursing and healthcare.
   • The agent consists of four specialized agents:
     • Social Media Manager Agent
     • Content Writer Agent
     • Image Generator Agent
     • Facebook Manager Agent
   • Reduces manual workload by 80-90%.
   • Human review step in Notion for ad approval.
   • Make automation posts approved content to Facebook.
   • Can analyze ad performance and provide recommendations.
   • Key takeaway: Human review step is crucial for mission-critical agents.
2. DNS Support Agent for Ali EII
   • Ali EII is an AI-powered COO automation platform.
   • The agent acts as a virtual network engineer.
   • Trained on major DNS registers’ documentation.
   • Can analyze DNS setup and provide step-by-step fix instructions.
   • Reduces customer support workload by over 80%.
   • Key takeaway: Combining knowledge and actions achieves higher performance.
   • Cost: About $0.08 per request.
3. Questionnaire Agent for a Sustainability Consultancy
   • The consultancy helps companies manage carbon emissions and reporting.
   • The agent processes questionnaires and applies scoring rubrics.
   • Reduces process time from 3 days to about 20 minutes.
   • Key takeaway: Running an agent is cheaper than manual processes.
   • Cost: Between $1and$2 per process.
4. Home Repair Estimate Agent for Jack
   • Jack is a startup focused on simplifying home maintenance and repair.
   • The agent engages homeowners and retrieves pricing data for estimates.
   • Reduces administrative overhead by up to 90%.
   • Key takeaway: Ensure the agent gathers all necessary information.
   • Cost: About $0.10 per estimate.
5. CO Reporter Agency for Ali AI
   • Generates reports from data sources like Google Analytics.
   • Consists of a COO agent connected to three other agents.
   • Reports are generated in minutes with fewer errors.
   • Can create graphs and visual comparisons.
   • Generates PDF and Excel reports with charts.
   • Key takeaway: Agents need to be adaptable based on feedback.
   • Cost: About $0.12 per report.

Conclusion

• Most agents are specialized for specific customers.
• Horizontal agents can be reused across multiple businesses.
• Insights from horizontal agents can lead to scalable vertical solutions.

Additional Notes

• The video demonstrates the practical application of AI agents in various industries.
• The agents are designed to automate complex processes and provide significant time and cost savings.
• The importance of human review and feedback in the deployment of AI agents is emphasized.
• The potential for productizing and scaling AI solutions is discussed.