[
  {
    "i": 0,
    "speaker": "Speaker 1",
    "text": "To describe uh what we do in a nutshell, we take a history snapshot of the market and try uh to predict the price in the future and then try to make the best possible trading decision."
  },
  {
    "i": 1,
    "speaker": "Speaker 1",
    "text": "I find it fascinating because >> uh >> the market is such a dynamic and complicated and noisy environment that uh we have to understand the problem and the solution from the very first principle and to a very great depth if you want to make a trade in different horizon."
  },
  {
    "i": 2,
    "speaker": "Speaker 1",
    "text": "They are completely different approaches and they're completely uh different problems even though they look like they're so similar. We have a large amount of data in the market that uh we can test a lot of complicated solutions on. One of the most exciting thing I find about my"
  },
  {
    "i": 3,
    "speaker": "Speaker 1",
    "text": "work is we have a lot of leeway on what kind of technique we can apply. We survey a range of methods >> uh >> and then we apply whatever works for a problem."
  },
  {
    "i": 4,
    "speaker": "Speaker 1",
    "text": "If you can prove something is working better in a statistically uh rigorous way, then it can be rolled out trading live within a matter of few days. The thing with research is for the most time you're wrong. Maybe nine out of 10 attempts uh are failure if not more."
  },
  {
    "i": 5,
    "speaker": "Speaker 1",
    "text": "Every failure is a learning experience when the theoretical hypothesis uh clicks with what we see in live trading. When you put the pieces together, the satisfaction is beyond expression. uh"
  }
]