In this demo, we introduce the tweet-based newsfeed summary service, called iApollo, running on a named data network (NDN) stack. This novel application provides a customized newsfeed service to individual readers based on their interests. Data sampling is essential in iApollo because of the large volume of tweets. Espresso, the automatic naming agent, translates this sampling problem into the simple tree traversal by constructing a hierarchical namespace for the given set of tweets. Two types of tree traversals are introduced: a modified breadth-first-search (BFS) and depth-first-search (DFS) which result in generating headline and complementary news, respectively. This allows readers to quickly achieve the best semantic understanding of the news topic with the minimal number of data retrievals. The newsfeeds are transmitted over NDN because NDN caching can reduce the retrieval delay while releasing burden on the end server. Furthermore, Espresso is well-matched to NDN as it naturally resolves NDN's naming requirement. The application demonstrates not only the tweet-newsfeed service, but also the great synergy between Espresso and NDN. This work can be extended into a general software framework for content summarization.