Overview

Influencer marketing teams write the same kinds of email all day: meeting requests, negotiations, follow-ups and collaboration pitches. This tool turns a short brief (influencer, subject, what you want) into a personalised, professional email, then keeps a searchable record of every message sent.

How it works

1. Brief: in the React dashboard, the user picks an influencer, a subject, a category (meeting, negotiation, follow-up or collaboration) and a short description.
2. Generation: the Flask API asks OpenAI to write the email in a professional, concise voice, using the influencer's name.
3. Fallback: if the model is unavailable or no key is configured, a simple template is used instead, so the workflow never blocks.
4. Tracking: each email is stored in MongoDB with a generated task ID (TASK_YYYYMMDD_NNN), its category and a timestamp.
5. Views: sent-email list, timeline, calendar and analytics pages, filterable by influencer, category and date range.

Engineering decisions

• Structured requests and responses: every call uses typed request and response types (for example send_email and send_email_response), so the backend can be driven by the UI or by another agent.
• Degrade gracefully: an outreach tool that stops working when the LLM does is worse than no tool, so there is always a fallback path.

Tech stack

Python · Flask · OpenAI · MongoDB · React · REST API

Python

OpenAI

MongoDB

React

View project