30 Fastest Growing Companies in Asia 2019 30 Fastest Growing Companies in Asia 2019 | Page 55

a student. I excelled at school, jumped a
grade when I was in elementary school,
started in the computer programming
experimental program when I was
eight years old, got Academic Olympic
prizes, and was admitted into one of
the top China Universities by merit
without taking the college entrance
exam.” Because of those advantages,
he was able to recognize the need to
create a program to aid students who
are deficient in areas and possibly not
know it.
“I spent one-fifth of the time learning
compared to the other students and
got better results. Not because I was
smarter, but because I mastered the
methods of learning and took advantage
of the great teachers I was exposed to.
I found out that a master teacher is 10
times or even 100 times better than a
mediocre teacher, but such a resource is
just so rare.”
The idea was there, but how to put
it all together was still eluding him.
“Later, when I found out that Artificial
Intelligence can mimic a master teacher
and clone them to deliver personalized
learning at scale and can potentially
deliver such high-quality learning
impact to every student, no matter
if they are in a second-tier or even a
third or fourth-tier city, no matter
if the parents are rich or poor. Such
technology and algorithm can change
the game of education and make
high-quality education more fair and
accessible.”
Mr. Li explains that Squirrel AI Learning
believes that the traditional lecture-
centered factory model of learning
where a full classroom of different
students follow the same path at the
same pace where the instructor is
proceeding without individualized
attention to each individualized
student is by design inefficient. “We
built a mastery-based, fully personalized
tutoring model in which students can
receive immediate feedback and fully
adaptive instruction. We created an
artificial-intelligence-powered virtual
tutor engine, which can diagnose and
remediate each individual student’s
knowledge gaps in real-time while a
student is learning and practicing.”
Elements: That Make Squirrel
AI Learning Outshine
Competition
Whenever the student struggles or does
not meet a proficiency requirement of
a particular knowledge point, Squirrel
AI Learning’s adaptive engine would
introduce relevant instructional
content (such as videos, concept
reviews, instructions on solving
exemplar questions, mistake analysis,
etc.) and assessment content (such as
pre-test, post-test, practice questions,
quizzes, etc.) remediating the
knowledge gap and helps the student to
complete an optimized journey through
a personalized pathway towards the
mastery of materials.
“Even though a human tutor can
intuitively perform such tasks, we believe
we can only implement such a superior
pedagogical model in scale and with
consistent quality through AI-powered
adaptive platform.”
Through machine learning algorithms
and data analytics techniques, adaptive
learning systems offer learning
experience personalized to students’
characteristics and abilities. The intent
is to determine what a student really
knows and to accurately, logically move
the student through a sequential path
to prescribed learning outcomes and
skill mastery. Many learning products
with adaptive features have been
developed, such as Cognitive Tutor,
Knewton, RealizeIt, i-Ready, DreamBox
Learning, ALEKS. Such systems
constantly collect and analyze students’
learning and behavior data and update
learner profiles. As students spend
more time in it, the system better
knows their ability and can personalize
the course to best fit their talents.
“In October 2017, Squirrel AI Learning
successfully held the first man-machine
learning competition in the AI field
in Asia-Pacific region; the Squirrel AI
Learning teaching robot outdid a senior
teacher with 17 years of experience –
scoring nine points.”
The event was widely covered by major
media outlets such as CCTV, U.S. CNBC,
Japan’s NHK Channel, Forbes, FORTUNE
China, Chinese Business News, etc.
All Set to Expand into
International Territory
“We believe the adaptive technology
can be applied as a general method for
a wide area of learning scenarios. Even
within China, there are great varieties of
content, curriculum and requirements
across the many provinces and cities
we are operating in. As each student
is by nature different in character,
knowledge, and capacity, we are solving
the generic problem of dealing with
these differences using the adaptive
approach. We have already been
approached by several US, European,
and Asian education institutions who
are interested in adopting our platform
in their countries. When we started
our program, we also leveraged the
experience and lessons learned from
the US and Europe. Our approach to
expanding into future international
markets is to take advantage of our
experience in localizing our platform by
integrating local content to optimize for
the local education policy, local teaching
and testing requirements as well as
regional student characteristics.”
“We are planning to expand into the
international market within three years
and if the conditions are appropriate,
we might even consider accelerating the
timeline. First, we need to excel in the
Chinese market, which is our foundation
and we need to establish enough
operation and data size to optimize our
core platform. Once we gain enough
experience and operational expertise,
we will be able to rapidly replicate our
success. We will consider all options for
our international presence, including
direct operation, joint venture, platform
collaboration, direct or indirect
investments, and the like.”
SR