Key Concepts
AI Agents, AI Agent Economics, Cost Tracking, Value-Based Pricing, Trucking Industry, AI-powered Dispatch, Risk Assessment, Home Hardening, Insurance Underwriting, Sales Cycle Hacking in Regulated Industries.
Paid AI: AI Agent Economics and Value-Based Pricing
Overview
Paid AI, founded in January, focuses on tracking AI agent costs and facilitating monetization for companies building AI agents. They help companies understand the economics of running AI agents, which differ significantly from traditional SaaS models due to the cost of inference and tokens.
The Agent Market
The AI agent market is nascent. While anyone can build an agent, few can create one that solves a problem customers are willing to pay for. Agents are being adopted rapidly in non-tech sectors like plumbing, architecture, and mortgages.
Economic Cycles and Agent Adoption
Unlike traditional software adoption where software startups buy from each other, agents are seeing direct adoption by "real business companies" in sectors like freight forwarding, supply chain, and insurance. This is driven by an aging workforce and the efficiency of agents compared to Business Process Outsourcing (BPO).
Cost Structure of Agent Companies
Traditional SaaS companies have gross margins of 70-80%. Agent companies, due to inference and token costs, have margins of 40-60%. This is highly variable depending on customer usage. Verbose customers or complex tasks consume more tokens, increasing costs.
Multiple AI Model Providers
While some systems use multiple AI model providers (e.g., GPT, Claude, Gemini), this is primarily for behavioral differences rather than economic arbitrage. Each model has unique tendencies, making direct swaps challenging.
Paid AI's Approach to Cost Tracking
Paid AI helps companies understand their costs and the value their agents provide. Pricing has three components: cost, equivalent service cost (human), and competition. Paid AI helps companies understand what the agent is doing, allowing them to pitch value to customers effectively.
Granular Cost Tracking with Open Telemetry
Paid AI uses Open Telemetry to capture the runtime code of agents, providing a detailed log of every activity. This allows for granular tracking of costs down to the cent level.
Value-Based Pricing
Paid AI advocates for flexible pricing based on customer-perceived value rather than a consistent price sheet. This means aligning the price with the value the customer receives from specific agent activities.
Example: Car Dealership Customer Service Agents
A car dealership using customer service agents sees different conversion rates in different locations (South Dubai vs. North Dubai). Driving a lead to a South Dubai dealer is more valuable due to higher conversion rates, justifying a higher price.
Shift to Pay-for-Demonstrated-Value
The discussion suggests a shift from per-seat pricing to pay-for-demonstrated-value, where customers only pay when an agent successfully completes a transaction.
Challenges and Future Interactions
Implementing value-based pricing requires detailed information and flexible systems. Paid AI aims to provide this flexibility, enabling companies to charge based on the value delivered to each customer.
Services and Automation
As agentic services replace traditional services, the focus shifts to higher-level services like marketing programs and upselling. The goal is to automate discriminatory pricing based on location and other factors.
Key Quote
"You shouldn't price in a way that is simple to you. You should price in a way that is consumable by your customer, in a way that your customer perceives they're getting value." - Manny Medina, Co-founder and CEO of Paid AI
Conclusion
Paid AI addresses a critical need in the AI agent economy by providing tools to track costs and implement value-based pricing. This approach puts pressure on AI agent companies to build high-quality products that deliver tangible value.
iTrucker: AI-Powered Dispatch in the Trucking Industry
Overview
iTrucker, based in Pittsburgh, uses AI agents to improve efficiency in the trucking industry. The company focuses on automating communication and coordination between drivers, dispatchers, customers, and brokers.
The Problem: Inefficient Communication
The trucking industry relies heavily on phone calls for dispatch, freight management, and communication. This is inefficient and creates challenges for drivers and dispatchers.
Camilo Ramirez's Background
Camilo Ramirez, co-founder and CEO of iTrucker, is a former truck driver and fleet owner. This experience gives him unique insights into the pain points of the industry.
iTrucker's Solution: AI Agents for Communication
iTrucker deploys AI agents to automate communication, book loads, and forward information. These agents can interact with drivers, customers, and brokers via voice or text.
AI Agent as a Co-Pilot
The AI agents act as co-pilots, assisting dispatchers in managing their fleets. Dispatchers have a dashboard to monitor AI activity, approve suggestions, and intervene when necessary.
Example: Idle Truck Notification
If a truck is idling for an extended period, the AI notifies the dispatcher and suggests finding a load. The dispatcher approves the suggestion, and the AI books the load and notifies the driver.
Data Sources: ELD Integration
iTrucker integrates with Electronic Logging Devices (ELDs) to access real-time data on truck location, speed, fuel consumption, and hours of service.
Market Size and Target Customer
iTrucker targets mid-size fleets with 3 to 100 trucks. There are approximately 300,000 such fleets in the US, providing a large market opportunity.
Enterprise Accounts
iTrucker is seeing opportunities with larger enterprise accounts, but these require a different sales approach and may involve custom software development.
Sales Cycle and Go-to-Market Strategy
The current go-to-market strategy involves founder-led sales, targeting owners of small trucking companies. Enterprise sales require a longer sales cycle, more resources, and compliance with standards like SOC 2.
Advice for Handling Enterprise Customers
Jason Calacanis advises iTrucker to study their customer base and decide whether to focus on small customers or pursue larger enterprise deals. He suggests charging a significant activation fee for custom software development and embedding a product manager at the customer's company.
Key Quote
"Don't be scared of these customers... You got to kind of let them have skin in the game." - Jason Calacanis
Conclusion
iTrucker is using AI to address inefficiencies in the trucking industry. By automating communication and providing real-time insights, iTrucker helps trucking companies improve their operations and reduce costs.
10X AI: Risk Assessment and Home Hardening for Insurance Underwriting
Overview
10X AI, led by CEO and co-founder Elise Myn, uses AI and computer vision to assess risk at the individual property level. The company aims to help homeowners protect their homes from extreme weather and enable insurance companies to underwrite risk more accurately.
The Problem: Inaccurate Risk Assessment
Insurance companies typically use regional models to assess risk, leading to inaccurate assessments and unfair pricing. Manual inspections are time-consuming and inefficient.
10X AI's Solution: AI-Powered Risk Assessment
10X AI uses computer vision and AI models to analyze images of properties, providing a granular view of risk. Homeowners can collect images using a mobile app, or drones can be used for verification.
Efficiency and Accuracy
10X AI's process is 80% more efficient than manual inspections. The AI models can identify specific risk factors, such as fence material, window type, and vent mesh.
Regulatory Changes in California
California recently approved catastrophe models that allow insurers to incorporate specific data about homes into their risk assessments.
Risk Factors and Mitigation
10X AI identifies risk factors and quantifies their impact on risk reduction. For example, the right type of roof can reduce risk by six times.
Target Customer and Users
Insurance carriers are the primary customers, while homeowners are the users of the software. Homeowners can access the software through their insurance provider and receive discounts for taking risk-reducing actions.
Community-Wide Approach to Risk Reduction
10X AI is seeing interest from communities and HOAs in taking a community-wide approach to risk reduction. Hardening multiple homes in a neighborhood can improve the survivability of the entire neighborhood.
Home Hardening and Retrofitting
There are many things homeowners can do to harden their homes against fire and flood. Investing in the right retrofits can save at least six times the recovery cost.
Sales Cycle Hacking in Regulated Industries
Elise Myn seeks advice on hacking sales cycles in regulated industries. Jason Calacanis suggests targeting high-end homeowners and creating content (e.g., YouTube channel) to raise awareness and establish expertise.
Key Quote
"At the very minimum $1 invested in the right retrofits or controls on a property saves at least at a minimum six in recovery cost." - Elise Myn, CEO and co-founder of 10X AI
Conclusion
10X AI is addressing a critical need in the insurance industry by providing accurate and granular risk assessments. By empowering homeowners to take risk-reducing actions, 10X AI helps protect homes and maintain insurability.
AI summaries can miss context or contain errors. Check important details against the original video.