Elasticsearch ngram index returns nothing

Question

I'm attempting to build a custom analyzer using nGram and apparently it's working ok, I just can't query it for some reason. I'm using `elasticsearch-model in Ruby

Here is how the index is defined:

  include Elasticsearch::Model

  index_name "stemmed_videos"

  settings index: { number_of_shards: 5 },
    analysis: {
      analyzer: {
        video_analyzer: {
          tokenizer: :stemmer,
          filter: [
            "lowercase"
          ]
        },
        standard_lowercase: {
          tokenizer: :standard,
          filter: [
            "lowercase"
          ]
        }
      },
      tokenizer: {
        stemmer: {
          type: "nGram",
          min_gram: 2,
          max_gram: 10,
          token_chars: [
            "letter",
            "digit",
            "symbol"
          ]
        }
      }
    } do
    mappings do
      indexes :title, type: 'string', analyzer: 'video_analyzer'
      indexes :description, type: 'string', analyzer: 'standard_lowercase'
    end
  end

  def as_indexed_json(options = {})
    as_json(only: [:title, :description])
  end

I've attempted to take one of the strings I'm trying to index and run it through "http://localhost:9200/stemmed_videos/_analyze?pretty=1&analyzer=video_analyzer&text=indiana_jones_4-tlr3_h640w.mov" and it's apparently doing the right thing. But then, the only way I have to make a generic query is by adding wildcards, which is not what I'm expecting.

[8] pry(main)> Video.__elasticsearch__.search('*ind*').results.total
=> 4
[9] pry(main)> Video.__elasticsearch__.search('ind').results.total
=> 0

(4 is the right number of results in my test data). What I'd love to accomplish is to get the right results without the wildcards because with what I have now I'd need to take the query string and add the wildcards in the code, which honestly is rather bad. How can I accomplish this?

Thanks in advance.