Clashpoint

The problem it addresses

Sales coaching rests on two activities: reviewing calls that have already taken place and rehearsing conversations before they happen. Both normally involve another person: someone who listens to a call, judges how it went and turns that into specific feedback, or someone who plays the other party in a practice conversation. Clashpoint aims to support both with AI, which analyses and scores calls, generates feedback for the rep and acts as a practice partner.

What it does

Clashpoint covers both call review and practice in one platform. Sales calls are transcribed, analysed with AI and scored, and the rep receives feedback based on that analysis. For practice, reps hold roleplay conversations in which the other party is an AI, so they can rehearse a conversation before having it for real. The product is built around four functions:

• Call analysis: transcription and AI analysis of sales calls
• Scoring: scores for analysed calls
• Feedback: comments for the rep, based on the analysis
• AI roleplay: practice conversations in which the other party is an AI

How it is built

Clashpoint consists of an API built with FastAPI, a web application built with Next.js and React, and background workers that handle transcription, scoring and feedback. Data is stored in PostgreSQL, and the platform's AI features use large language models (LLMs). The whole product is developed in-house at KS Venture Group, from the API and database to the web application and the AI features.

Status

Clashpoint is in development and is one of KS Venture Group's active portfolio projects. Work on it began in 2024 and is done in-house. The product's scope covers the functions described above: call analysis, scoring, feedback and AI roleplay. This entry will be updated as the product develops.