PrimRose

Personal Relationship Managing Strategic Opportunity Engine It's an engine for software engineers to to tap into their network to find and validate ideas for startups and SaaS businesses.

Comment

PrimRose aka PRMSOE: Strategic Opportunity Engine (prmsoe.com)

Inspiration

I wanted to build a SaaS but I wasn't sure what to build or how to get started. I watched a lot of videos and read a lot of books that say to start with the customer first instead of building first. So I wanted to build a tool that can assist me in accessing, analyzing, and reaching out to people in my network to find and validate ideas.

What it does

PRMSOE (pronounced Prim-So) is a "Validation Engine" that turns your LinkedIn network into a laboratory for product-market fit.

  • Intelligent Ingest: Seamlessly parses your LinkedIn Connections.csv export and maps it to a secure, private database in Supabase.
  • Founder-to-Market Matching: Leverages the advanced reasoning of Gemini to match your personal skills (parsed from your resume) against the professional roles of your contacts to identify the highest-probability validation paths.
  • Strategic Research: Integrates the You.com API to research the latest technical hurdles at your contacts' companies, providing "Earned Insights" before you even reach out.
  • Low-Friction Outreach: Generates peer-to-peer message drafts designed specifically for problem discovery, optimized for high response rates from industry professionals.

How we built it

We prioritized a high-performance, AI-native stack to handle the next-gen reasoning of Gemini 3:

  • Frontend: Lovable (React) for a sleek, glass-morphism dashboard and rapid UI iteration.
  • Backend: FastAPI hosted on Modal for scalable, serverless AI orchestration and concurrent processing.
  • Intelligence: Geminifor deep strategic reasoning and intent analysis.
  • Search: You.com API for real-time web grounding and technical signal detection.
  • Database: Supabase with Row Level Security (RLS) to ensure user data remains private and secure.

Challenges we ran into

The primary challenge was Inference Depth. It’s easy to summarize a profile, but it’s hard to prompt an AI to find a "validatable" business problem. We utilized Gemini 3’s advanced reasoning to move beyond simple summarization into "discovery partnership." Additionally, handling the inconsistent data formatting within LinkedIn's CSV exports required a robust normalization layer in our FastAPI backend to ensure the AI always had clean, structured context.

Accomplishments that we're proud of

  • The "Validation" Pivot: Successfully moving beyond generic outreach to generate insights that actually spark discovery conversations using Gemini 3's sophisticated tone control.
  • Performance: By utilizing Modal for concurrent API calls, we reduced the "Time to Strategy" from 15 seconds down to under 3 seconds.
  • Privacy-First Architecture: Implementing a robust Supabase RLS structure that guarantees a user's network data is never shared or used for training.

What we learned

How to use the tools provided. We learned about how to use and deploy to render. We learned more about You.com, AGI, and Composio.

What's next for prmsoe.com

  • Multi-Source Ingest: Expanding beyond LinkedIn to include Google Contacts and Google Maps Timeline data to reconstruct the "where" and "when" of every connection.
  • Autonomous Watchtower: A proactive agent that monitors your network and pings you when a contact's company context changes in a way that aligns with your founder expertise.
  • Direct Outreach Integrations: Building native connectors for Email and professional messaging platforms to close the loop from "Inference" to "Conversation."

Built With

Share this project:

Updates