Cross-Platform, Socially-Aware Media Recommender Powered by AI
Our media lives are fragmented. Between streaming services, social DMs, and personal note apps, people are constantly managing what to watch, read, and revisit—but nothing brings it all together. Recommendation engines offer suggestions, but they don’t reflect the emotional or social context behind what we actually love.
I built the concept for Streamlink to explore how AI can reimagine recommendation systems to feel more human, connected, and helpful—mirroring the way we organically discover media through our friends, moods, and memory.
Today’s recommendation systems are:
Siloed by platform (Netflix, Prime, Hulu…)
Out of sync with real behavior (rewatches on other platforms aren’t counted)
Socially blind (no awareness of what your friends recommended)
Hard to manage (most people keep lists in their Notes app or texts)
We’re left juggling watchlists, remembering what someone suggested three months ago, or wondering what to watch next without helpful context.
There’s an unmet need for a socially intelligent, AI-powered concierge—one that can:
Parse and organize recs from natural text (Notes apps, messages)
Track availability across platforms and notify you when things change
Understand your media taste and that of your trusted friends
Answer questions like: “What can I watch tonight that’s short, nostalgic, and available now?”
A personalized media companion with four core AI-driven functions:
Smart Rec Organizer: Imports messy text and turns it into structured lists (title, genre, platform, source)
Conversational Discovery Agent: A queryable AI that helps answer “what should I watch/read?”
Friend Graph for Taste: Track who recommended what and match your preferences
Streaming Tracker: Alerts for expiring or newly added titles based on your saved lists
OpenAI GPT-4 / Claude 3 – NLP for parsing and conversation
JustWatch API + LangChain – Streaming availability engine
Airtable or Notion AI – Dynamic watchlist UI
Zapier + Replit – Automations (alerts, reminders)
Streamlit or Figma – Demo-ready interface
This project highlights my interest in solving real user problems at the intersection of content, community, and AI. It’s a passion-fueled exploration into the future of recommendations—rooted in trust, memory, and human connection rather than isolated algorithms.
As recommendation engines become more complex and content continues to splinter across platforms, we need tools that reflect how people actually consume media—socially, emotionally, and asynchronously. AI gives us the ability to close that gap.