A research engine for connected knowledge. Explore a company, a paper's citations,
or the hidden path between any two ideas across Wikipedia, research papers, finance, and biology
— every connection backed by real evidence.
Explore
→
Press Enter to search
Aurelius
FROM
TO
Ready.
Search
0 explored
Compare & explain
See how any two things connect: the strength of the link, the chain of
real relationships behind it, and how their prices move together.
vs
Coverage
Discover hidden connections
What is a hidden connection? A relationship that a
normal search can't show you because no direct link exists. Aurelius
walks everything one node touches, then everything those
touch, and keeps candidates with strong indirect evidence: many
shared neighbours, semantic similarity, co-movement, shared
dependencies, or common intermediaries. Each result explains why it
surfaced before you explore it.
Search any company to see its price, peers, supply chain, correlations, ownership and the news moving it — all in one place.
Give it a paper — by title, DOI, or arXiv id — to see what it cites, what cites it, and build an interactive citation graph you can expand.
Path & Details
Path
Scores
Log
Key
Path will appear here as Aurelius explores.
Current centre (being expanded)
Open, queued to explore
Closed, already explored
Target article
On current best path
f(n) = total estimated cost g(n) = hops from start h(n) = semantic distance to the target
Aurelius always expands the node with lowest f(n) next,
and clicking any node after a search expands its real neighbours.
Path Discovered!
—Hops
—Steps
—Time
Every link above is real, but the wording shown on a page isn't always
the link's title; for example, a page may display "iterative" while the link
actually points to "Iteration", or show "AI" linking to
"Artificial intelligence". Wikipedia links can be piped like that, so
the visible text and the article it leads to can differ.
AURELIUS
Graph intelligence engine
"
The universe is change, our life is what our thoughts make it.
Marcus Aurelius, Meditations
Named after the Stoic emperor who wrote that every idea connects to the next. Aurelius
makes those connections visible in any network of knowledge: pick two things, and it walks the real links
between them, ranking every step by meaning. It started on Wikipedia; today the same engine powers a
company research workspace, a paper citation explorer, biomedical networks and live news coverage,
and can surface connections that no direct link records at all.
How it works
Works best on desktop; the live graph and side panel are easiest to read on a larger screen.
Best-First Search
Instead of checking links at random, Aurelius always opens the article that
looks closest to the goal next, like always taking the turn that points toward where
you're headed. It's the same idea behind A*: chase the most promising lead first rather than
exploring everything.
Semantic Embeddings
Each article is turned into a string of numbers that captures its
meaning. Related topics get similar numbers, so "Jazz" sits near "Saxophone" but far
from "Java". Comparing those numbers is how the search tells whether a link actually gets it
closer.
Search From Both Ends
Aurelius also gathers everything that links to your destination
ahead of time. When the forward search runs into one of them, the two sides meet in the middle,
reaching the answer far quicker than searching from the start alone; and every hop in the result
is a real, verified edge.
Breaking Out of Dead Ends
If the search keeps circling a cluster of similar-but-wrong pages, it notices
it has stopped making progress, casts a wider net, and favors more varied links to escape,
and it never invents a link that doesn't truly exist.
Hidden Connections
A search engine can only show you what's directly linked. Aurelius' most useful trick is
finding what isn't. It walks everything one thing touches, then everything those touch, and
keeps the candidates with strong indirect support: many shared neighbours, close meaning, price
co-movement, shared suppliers or investors, common intermediaries. Two companies that never mention
each other but depend on the same chip maker; a drug and a disease with no listed link but a chain of
shared biology between them. Every hidden connection comes with the evidence for why it
surfaced, so you're never asked to trust an unexplained edge.
What makes it different
Typical search
Aurelius
Treats every link as equal
Ranks every link by meaning
Searches blindly until it stumbles on the target
Aims at the goal from both ends
Built for one dataset
Same engine on Wikipedia, papers, finance, biology & news
Only finds what's directly linked
Discovers hidden connections with no direct link at all
One engine, many worlds
Every dataset here is a plug-in adapter speaking one protocol, so the search engine
never changes: Wikipedia and research papers are walked live; finance and news are ingested into a
persistent graph with learned structural embeddings on top. The finance graph links companies, ETFs,
sectors, executives and macro indicators through typed relationships (supply chains, ownership,
competition, real return correlations), and a standalone News Intelligence pipeline turns live
headlines into entities, relationships and evolving stories that any domain can query. Every path hop
and every discovered connection comes with the evidence for why it exists.
Built by
Murtaza Vali
CS student who got curious about A* pathfinding and ended up
building something with embeddings, a verified backlink graph, and a live force-directed visualization. This is
what happens when you keep adding "just one more thing."
Wikipedia, research papers, finance, or biology. Each has its own workspace — a company
profile, a paper's citations — plus the pathfinder that ties any two things together.
2
Dive into one thing, or connect two
Research a company (price, peers, supply chain, news) or a paper's citation tree; or type
a From and a To and watch Aurelius walk the real links between them, ranking every step by
meaning. It never invents a link.
3
Explore the graph
Click any node to expand its real neighbours, focus on its neighbourhood, filter the view,
and open Discover to surface connections with no direct link at all.
Tip: in Finance, hit Research a company and search a ticker like NVDA — you'll see
its peers, suppliers, correlations and news before you ever touch the graph.