Polaris

Hi, we are Polaris, and today we are trying to make academic research a more pleasurable experience.

Inspiration is one of the biggest challenges we face when creating application designs, doing project ideation, and conducting academic research.

Polaris solves this abstract issue by connecting the nodes to create a network for inspiration.

So, how does Polaris solve this?

One of the most important components within research is finding related papers and branching off further keywords for citation.

When a keyword is given, Polaris queries OpenAlex, an academic database pulling both review papers and top-cited articles. It then feeds the topics, abstracts, and subfield hierarchies from those papers into Gemini, which organizes them into a structured knowledge graph. Not from its own imagination, but grounded in real literature.

Every node in the graph traces back to published research.

When you expand a node, the system sends its full ancestry chain (parent, grandparent) to the LLM, giving our research agent a more refined output, compared go generic concepts.

And when you click "Deep" on any node, Polaris pulls the top 10 cited papers from OpenAlex, loads them into context, and lets you have a grounded conversation with those papers, with inline citations.

Polaris enables us to gain deeper insights while conducting research, where can connect two dots that could have never been connected.