PitchLab

Inspiration

Around 79% of sales pitches fail, with every bad pitch costing companies millions in lost revenue. However, there isn't enough practice. Managers are too busy, role-play mocks are not entirely accurate to the real thing, and you only get one quick shot with a real prospect.

What if there was a product that could mimic the exact prospect you'll be speaking too?

What it does

PitchLab generates a simulated prospect for you to practice sales pitches with based on your learning goals, company-specific data, and industry research. We use large language models (LLM) and AI agents to act autonomously on a company's messy, real-world data. After analyzing company sales data, we generate a simulated pitch call environment with an AI prospect. This AI prospect speaks to you in real time, and is catered based on the prospective company, the sales rep's learning goals, and company product suite

How we built it

We used NextJS, TypeScript, TailwindCSS, ReactJS for the front-end development. We used Go for the backend. Claude, ElevenLabs, and AI agents (similar to what Rox offers) create a perfect user prospect that sales reps can practice with.

Here are the presentation slides! https://docs.google.com/presentation/d/1nO3dRgHFU7kB5YlKxWr52657JoPnywPV00_sTdDzK0Y/edit?usp=sharing

Built With

• claude
• elevenlabs
• go
• nextjs
• react
• rox
• tailwindcss
• typescript