Hashnode is an isomorphic app, which means the views can be rendered on both client and server. Basically, whenever someone requests a page from Hashnode (irrespective of who is requesting - bot or real user) they will see a server rendered version. This is easily achievable because we use Node.js on server and React components can be rendered on server via Node.
Once a page loads, all further interactions, page loads etc use client side rendering with history.pushState. So, this way you get best of both worlds.
To sum up, we don't use any extra tech/solution (like prerender.io) to make Hashnode crawlable. We support both server and client side rendering and therefore don't face any SEO problems.