Skip to content

nmusoko/ai-sales-copilot

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🤖 AI Sales Copilot

LangChain · Whisper · Streamlit · Python


📌 Overview

Sales representatives in small and medium businesses (SMBs) spend 10–20 minutes after every client call summarizing discussions, drafting follow-up emails, and updating CRMs manually.
AI Sales Copilot automates this process using generative AI — it listens, understands, and writes like a real sales assistant.


🎯 Problem → Solution → Impact

Problem
Post-call administrative tasks are repetitive, time-consuming, and error-prone.
Sales reps lose focus and valuable time that could be spent closing deals.

Solution
AI Sales Copilot automatically:

  1. Transcribes sales calls (via Whisper)
  2. Summarizes discussions and detects sentiment
  3. Generates a professional follow-up email in the rep’s tone
  4. Logs key action items and next steps in a structured CRM format

Impact

  • Saves ~10 minutes per call
  • Ensures consistent, high-quality communication
  • Keeps CRM data accurate and up-to-date

🧠 Key Features (MVP)

✅ Upload audio (.wav) or text transcript
✅ Automatic call transcription (Whisper)
✅ Smart summarization & sentiment analysis (LangChain + LLM)
✅ AI-generated follow-up email (tone-matched)
✅ Extracted “next steps” and action items
✅ Save/export results to a local SQLite CRM
✅ Streamlit dashboard for interactive use


📊 Success Metrics

Metric Target
⏱ Time saved per call ≥ 10 min
✅ JSON valid outputs ≥ 95 %
📨 Email quality rating (1-5) ≥ 4.0
⚡ Latency (audio → email) ≤ 2 min
🎯 Data accuracy (facts retained) ≥ 90 %

👤 User Persona

Role: Sales Account Executive (SMB)
Needs: Faster follow-ups, less admin time, consistent emails

User Stories

  1. Upload a sales call and receive a concise summary in < 30 s.
  2. Automatically generate a ready-to-send follow-up email.
  3. Extract next steps and deadlines for CRM logging.
  4. Download or export results as JSON or Markdown.

⚙️ Tech Stack

  • LLM & Orchestration: LangChain · IBM watsonx / OpenAI GPT
  • Speech-to-Text: Whisper
  • UI: Streamlit
  • Data Storage: SQLite · pandas · Chroma (optional memory)
  • Validation: Pydantic · JSON Schema
  • Environment: Python 3.10+

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published