AI Email Generator using modern IDE AI tools

---

Job Title: Simple Web Tool for AI Email Drafting with "Save to Drafts" (VC Outreach) Project Goal: Create a user-friendly web tool specifically for our Executive Assistant (EA). The tool will streamline the process of initiating outreach to VCs about relevant vendor offers for their portfolio companies. It will use AI (LLM) to analyze potential fits and draft the email, allowing the EA to review and save the approved draft directly to their Gmail drafts folder. Simplicity for the EA is paramount. User Workflow: EA accesses the tool via a web URL and connects their Google Account (one-time authorization). EA uploads the Portfolio Company List: A file (.csv or .txt) containing domain names (approx. 10k-20k lines). EA uploads the VC/PE List: A file (.csv or .txt) containing VC/PE firm names and primary contact email addresses. EA inputs Vendor Details via alink that we give with deal description. The tool processes the portfolio list against the vendor profile using an LLM (this may take a few moments; displaying progress is a plus). The tool displays: A list of portfolio companies identified by the AI as potential fits. The generated Subject line for the email. The generated Email Body (personalized for the selected VC contact, highlighting the vendor, offer, links, and the identified portfolio companies). EA reviews the generated text on the screen. If satisfied, EA clicks "Save to My Gmail Drafts". The tool uses the Gmail API to create a new draft email in the EA's connected Gmail account containing the generated Subject and Body. The tool displays a confirmation message (e.g., "Saved to your Gmail Drafts!") or an error message. EA can then go to their Gmail drafts folder to make final edits (if desired) and send the email manually. Technical Implementation: Platform: Simple, deployable web framework (e.g., Python + Streamlit, Flask, or similar). Prioritize ease of deployment and maintenance on standard cloud platforms (or potentially low-code solutions if they support the requirements). File Upload: Must handle uploading and parsing reasonably large .csv or .txt files (up to ~20k lines for portfolio, smaller list for VCs). Define expected format (e.g., one domain per line, or CSV columns). LLM Integration: Use a pre-configured LLM API (e.g., Google Gemini). API Key via secure environment variable. Consider efficiency for analyzing potentially thousands of companies (e.g., handling API rate limits, perhaps batching if possible, clear progress indication to the user). Google Workspace Integration (Crucial): Implement secure OAuth 2.0 flow for the EA to authorize the application to create drafts on their behalf. Must use appropriate scopes (specifically gmail.compose or gmail.modify to create drafts). Use the Gmail API Users.drafts: create method to save the generated content to the user's drafts. Handle token storage/refresh securely (leverage reliable libraries). User Interface: Clean, simple, intuitive for a non-technical user. Must include file upload widgets, dropdown for VC selection, text inputs for vendor details, clear display area for results/draft, the "Save to Drafts" button, and status/confirmation messages. EXPLICITLY OUT OF SCOPE: NO database for persistent storage (state managed per session or via simple browser mechanisms if needed for OAuth flow). NO automated sending of emails. NO scheduling or sequencing features. NO complex user account management beyond Google OAuth connection. Developer/Builder Profile: Experience building functional web applications with a simple UI. Must have proven experience with Google OAuth 2.0 implementation and the Gmail API, specifically creating drafts. Comfortable integrating third-party LLM APIs. Experience handling file uploads and basic data parsing in the chosen framework. Ability to build and deploy quickly, focusing on core functionality. Deliverables: A working, deployed web tool accessible via URL. Clear, concise instructions for the EA, including the one-time Google account connection process and how to use the tool day-to-day. Source code and deployment instructions/configuration details.

Keyword: marketing