I love exploring. I love climbing. I love setting out on a new path in the Colorado mountains, soloing 14ers, skiing down couloirs or climbing up ice waterfalls. The journey, not destination (as they say) is the goal.






Research is the same way. A curious mind never arrives at the end of the trail. There are endless paths and branches to explore. Yet we’ve always been confined to the 8x11 boxes of text, converting them from atoms to bytes and back, adding search and now Chatbots to try to generate new synapses. It’s not a intellectually stimulating way to learn.
I built Synaptica to solve my own problem. I wanted a better way to navigate AI research than reading isolated papers or scrolling through academic archives.
What It Does
Synaptica maps relationships between AI research concepts. Start with any topic (transformers, diffusion models, reinforcement learning, etc…)and the tool generates a concise summary of key papers. It then identifies core concepts, which you can click to explore further. Each click builds a visual map of your research journey.
Why I Made It
Reading disconnected papers makes it hard to see the big picture. Traditional search engines don't show how ideas connect. Chatbots can't map complex relationships between concepts. I wanted something that would help me discover meaningful patterns across research areas.
How It Helps You
Visual Research Trails: The interactive node map shows you exactly how concepts connect. You see your entire research path and can backtrack or branch off at any point. This gives you a bird's-eye view of a research area impossible to get from reading papers sequentially
RAG-Powered Exploration: The tool pulls information directly from a curated database of AI research papers. This means you get accurate, source-based information, not LLM hallucinations or innacurate citations from AI web searches.
For the nerds- Synaptica generates a dynamic knowledge graph with just a simple RAG vector store. No commercial graph databases and no complex logic. Just an elegant solution under few hundred lines of code.
Multi-LLM Support: Choose which AI model works for your needs. Some are faster, others more thorough. You control which engine powers your journey.
Concept Extraction: The system spots key technical terms in each summary. This surfaces entry points you might miss, leading to unexpected but relevant discoveries.
Research Synthesis: After exploring several connected topics, Synaptica can combine everything into one coherent summary. This turns your meandering exploration into a structured narrative.
Entity Explanations: Click any technical term for a plain-language explanation. No more switching to Google or ChatGPT to learn unfamiliar concepts within your research.
PDF Export: Save your entire research trail as a document.
Research Podcasts: Convert your findings into spoken audio in an educational podcast form. Listen to your research while commuting or exercising.
Try It
Synaptica is free- no login or sign up. Just start exploring.
Got An Idea?
Want to explore a specific kmnowledge domain? Case law? SEC filings? Let me know!
Want an account version of Synaptica where you can save research trails and even upload your own knowledge base to explore? Reach out!
Turn your company’s documentaion into an interactive guide? Get in touch!
Stay Curious My Friends!
-Nathan Staffel
nathan@revenantai.com