Problem / Background
Most networking tools are built for sales teams, not real humans with real relationships. People rely on scattered notes, inbox searches, and memory to remember who they should follow up with, what they talked about last time, and what’s happening in that person’s world. Important relationships go cold not because people don’t care, but because there’s no simple system that brings the right person and the right talking point to the surface at the right time.
Vision / Goals
Rolo is my attempt to build the AI-powered rolodex I wanted for myself: a lightweight, intelligent companion that keeps track of people, context, and momentum in relationships. The core goals:
- Make it easy to capture relationship context, not just contact info.
- Use AI and news data to surface relevant, timely reasons to reach out.
- Turn scattered meetings and notes into structured, searchable relationship history.
- Ship quickly as an iOS app with a real backend and real users—not just a slide deck.
The long-term vision is a personal networking OS that quietly works in the background, helping you connect with purpose instead of defaulting to “I should reach out sometime.”
My Role & Team
Rolo is a founder-led project that I’ve owned end-to-end. My responsibilities include:
- Defining the product vision, scope, and early feature set for the MVP.
- Designing the mobile experience (flows, UX, and visuals) for iOS in SwiftUI.
- Architecting and implementing the Supabase/PostgreSQL backend and schema.
- Building the AI and news-matching pipelines, including NER and prompt engineering.
- Running all early marketing: positioning, messaging, and content experiments.
In short: I’m acting as founder, product lead, engineer, and marketing team all in one.
Execution (Product + Tech)
The Rolo MVP combines mobile product design, a real backend, and AI-powered enrichment on top of news data.
Product & UX
- Designed the core flows in SwiftUI: contact list, contact detail, relationship context, and notes.
- Created a “pathways” mental model where contacts can be grouped by goals (e.g., job search, startup, creative projects).
- Designed meeting notes as short, structured entries that AI can later summarize and extract entities from.
- Planned a chatbot-style interface on each contact page so users can ask, “What should I talk to them about?” using all stored context.
Backend & Data Model
- Implemented the backend using Supabase with a PostgreSQL database.
- Designed schemas for contacts, meeting notes, entities (topics, companies, keywords), and news matches.
- Built secure API access for the iOS app to fetch and update data in real time.
- Set up the groundwork for role-based access and future multi-user features.
AI & News Matching
- Wrote a news scraping script to continuously pull headlines and article metadata from external sources.
- Built a named-entity recognition (NER) pipeline to extract people, companies, and topics from both contacts and news articles.
- Matched entities between a user’s rolodex and scraped news to generate curated, contact-specific news feeds and talking points.
- Used prompt engineering with LLMs to:
- Summarize meeting notes into relationship context.
- Generate suggested follow-up prompts.
- Propose conversation starters based on recent news.
Growth & Marketing
- Defined the positioning around “connect with purpose” and “networking, simplified.”
- Started experimenting with content and messaging on social channels to test resonance.
- Framed Rolo for students, early-career professionals, and builders as a way to keep track of their expanding networks.
Results / Impact
Rolo is still in active development, but even in the MVP phase it has already:
- Given me a single place to store relationship context across dozens of contacts.
- Shown that news-driven talking points can meaningfully change how and when I reach out to people.
- Validated that there’s demand among students and early-career professionals for a more human, AI-assisted networking tool.
More importantly, it proved that I can go from idea → design → backend → AI pipelines → working iOS app on my own, while also thinking about brand, positioning, and distribution.
Future Directions
Planned next steps for Rolo include:
- Refining the AI layer for better entity extraction and news relevance.
- Adding reminders and “nudges” based on relationship priority and time since last contact.
- Introducing collaborative features for small teams and founders.
- Building a cleaner onboarding that auto-enriches contacts from email, calendar, or LinkedIn (where allowed).
- Preparing a public beta with a tighter feedback loop from early users.
Takeaways / Reflection
Rolo has been a great sandbox for combining everything I care about: AI, product thinking, data, and real human relationships. Along the way, I’ve learned to:
- Scope and ship an ambitious product as a solo founder–builder.
- Design data models and pipelines that AI systems can reason over.
- Translate vague user needs (“I want to keep in touch better”) into concrete, shippable features.
- Balance building the product with learning how to tell its story and find the right audience.
Rolo is still early, but it’s one of the clearest examples of how I like to work: fast, experimental, and focused on tools that make people’s lives and relationships better.