1. Board Meeting Presentation
サービス・会社概要紹介
August 15th, 2013 - 3:30PM PDT
Treasure Data
Presented
by
Hironobu
Yoshikawa
–
CEO
Kazuki
Ohta
–
CTO
Rich
Ghiossi
–
VP,
MarkeIng
Keith
Goldstein
–
VP,
Sales
Kengo
Hirouchi
–
Director,
Japan
Ankush
Rustagi
–
Director,
MarkeIng
Founder & CTO
太田 一樹
<k@treasure-data.com>
www.treasuredata.com
Copyright
2013
2. 会社概要
チーム概要
2011年12月創業、米国カリフォルニア州。創業
者は日本人3人組。2013年12月現在、社員約
30名。
Hiro Yoshikawa – CEO
Open source business veteran
Kaz Ohta – CTO
Founder of world’s largest Hadoop Group
Jeff Yuan – Director, Engineering
LinkedIn, MIT / Michale Stonebraker Lab
ビッグデータの収集・保存・解析を一手に行える
クラウドサービスを提供。他のサービスと異なり、
数日で始められるのが特徴。
サービスコンセプト
• すぐに使い始められる
• クラウドサービスとしての提供を行う
• シンプルな機能セット、手厚いサポート
• “Trend Setting Products” in Data for 2014
(Database Trends and Applications)
• “5 Hot Big Data Startups”
(Enterprise Apps Today)
Keith Goldstein – VP Sales & BD
VP, Business Devt, Tibco and Talend
Rich Ghiossi – VP Marketing
VP Marketing, ParAccel and HP
投資家概要
Sierra Ventures – (Tim Guleri)
Leading venture capital firm in Big Data
Bill Tai
Renown investor, GP Charles River Ventures
Jerry Yang
Founder, Yahoo!
Yukihiro “Matz” Matusmoto
Creator, “Ruby” programming language
James Lindenbaum
Founder, Heroku
6. 6
各メディア / アナリストからの評価
“Treasure Data has taken a leadership position in providing
the first end-to-end public cloud-based big data analysis
service”
“A number of startups have begun to converge on the space
as well, including Treasure Data and BIME, which specifically
positions as cloud-based Big Data provider.”
“The question becomes, then, what role - if any - will the
public cloud play in helping enterprises turn Big Data into
actionable insights? Treasure Data believes it has an answer.”
“It’s only been six months since cloud data warehousing
company Treasure Data launched its services, but they’re
already reporting some impressive growth figures.”
11. 利用例: 14日間で月間600億インプレッションを裁くシステムを開発
1. ヨーロッパ最大のモバイル
アドエクスチェンジ
2. 2万5千以上のモバイルアプ
リから月間60億件以上のリ
クエストを裁く
3. サインアップから14日間、1
人のエンジニアによってシス
テムを完成させた
“Time is the most precious asset in our fast-moving
business, and Treasure Data saved us a lot of it.”
Julian Zehetmayr, CEO & Founder
12. Benefit – Reduce Cost and Complexity – Replace Hadoop
Before
1. Online Video Service serves
millions of users in 150
languages
2. In-house Hadoop cluster too
complex, costly and scaling
uncertain
After
3. Eliminated in-house Hadoop
cluster and redeployed
engineers on core businesses.
“Treasure Data has always given us thorough and timely support peppered
with insightful tips to make the best use of their service."
– Huy Nguyen, Software Engineer
18. Data Storage
Treasure Data Cloud
Default
(schema-‐less)
+me
v
1384160400
{“ip”:”135.52.211.23”,
“code”:”0”}
1384162200
{“ip”:”45.25.38.156”,
“code”:”-‐1”}
1384164000
{“ip”:”97.12.76.55”,
“code”:”99”}
•
•
Stored “schema-less” as JSON
– Schema can be applied/updated
AFTER storage
Compressed & columnar format
– For higher query performance
Schema
applied
~30%
Faster
+me
ip
:
string
code
:
int
1384160400
135.52.211.23
0
1384162200
45.25.38.156
-‐1
1384164000
97.12.76.55
•
Optimized for time-based filtering
•
Quickly scale-up processing power
99
www.treasuredata.com
Copyright
2013
– WITHOUT reloading/
redistributing the data
18
19. Board Meeting Presentation
August 15th, 2013 - 3:30PM PDT
新サービス & 新価格プラン
の発表
Presented
by
Hironobu
Yoshikawa
–
CEO
Kazuki
Ohta
–
CTO
Rich
Ghiossi
–
VP,
MarkeIng
Keith
Goldstein
–
VP,
Sales
Kengo
Hirouchi
–
Director,
Japan
Ankush
Rustagi
–
Director,
MarkeIng
www.treasuredata.com
Copyright
2013
20. 20
ビッグデータ活用:
7つのステージ
最適化
What s
a
trend?
Why?
アラート
Error?
ドリルダウン
Where
exactly?
アドホックレポート
Where?
定型レポート
レポーティング
予測分析
統計分析
データ解析
What s
the
best?
What
happened?
お客様の進化に合わせて、我々のサービスも進化を続ける。
25. 25
“ソリューションテンプレート”の提供
ソリューション
コンポーネント:
データ解析テンプレート
Treasure
Data
Service
データ収集テンプレート
- Treasure Data Service
- 構造化ログのテンプレート
- データ収集エージェント設定
ファイルテンプレート
- 設定済みBIレポーティング・
ダッシュボード
初期セットアップ期間内で、事前定義した解析ダッシュボードを提供
26. Board Meeting Presentation
Marketing Unified Analytics
Solution
August 15th, 2013 - 3:30PM PDT
Presented
by
Hironobu
Yoshikawa
–
CEO
Kazuki
Ohta
–
CTO
Rich
Ghiossi
–
VP,
MarkeIng
Keith
Goldstein
–
VP,
Sales
Kengo
Hirouchi
–
Director,
Japan
Ankush
Rustagi
–
Director,
MarkeIng
www.treasuredata.com
Copyright
2013
26
27. Business & Technical Problems
Marketing Tools data silos
–
•
Manual data integration
Difficult to merge other data
–
Data from online / offline systems
–
•
Website
C
Manual data pulling
–
Website
B
Sensor, CRM, ERP, Relational Data
Resource & time waste
–
Spend time pulling CSVs
–
Pull same data multiple times
–
Less time to focus on trends
–
MANUAL
PROCESS
•
Website
A
Hard to get cross-brand insight
Sensor
www.treasuredata.com
Copyright
2013
RDB
CRM
ERP
27
28. Marketing Tool Unified Analytics
Receive / Process
Monitor files,
process, and daily
sync to API!
Push to API
via Bulk
Import
Send to FTP
Files sent daily!
from SiteCatalyst!
FTP
Server
Store, Query, & Analyze
Automate queries
across multiple
profiles for KPIs!
BI
Connectivity
BI
Tableau, Metric
Insights, etc.
qp://qp.treasure-‐data.com/
www.treasuredata.com
Copyright
2013
28
29. Board Meeting Presentation
Gaming Analytics
Solution
August 15th, 2013 - 3:30PM PDT
Presented
by
Hironobu
Yoshikawa
–
CEO
Kazuki
Ohta
–
CTO
Rich
Ghiossi
–
VP,
MarkeIng
Keith
Goldstein
–
VP,
Sales
Kengo
Hirouchi
–
Director,
Japan
Ankush
Rustagi
–
Director,
MarkeIng
www.treasuredata.com
Copyright
2013
29
30. Goals & Solution
Analy+cs
Requirement
How
Treasure
Data
Delivers
Unify
AnalyIcs
in
One
LocaIon
Easily
and
automaIcally
load
data
to
cloud
DB
every
5
minutes
Drive
Cross-‐Game
Insights
Add
automated
queries
and
analyses
as
needed
Scale
and
adapt
to
new
tools
and
future
KPI
requirements
Flexible
database
and
data
collecIon
layers
Implement
quickly
with
no
upfront
costs
Provisioned
cloud
service
and
setup
within
weeks
or
IT
lag
Ime
Updates
and
changes
are
easy
and
take
hours,
not
weeks
or
months
Easy
to
use,
self-‐service
plasorm
and
robust
services
/
support
when
you
need
it
www.treasuredata.com
Copyright
2013
30
31. Treasure Data Gaming Solution
App
Developer
App
Developer
App
Developer
App
Developer
App
Developer
App
Developer
ApplicaIon
ApplicaIon
ApplicaIon
ApplicaIon
ApplicaIon
ApplicaIon
Log
Template
Data
Upload
Unified
Analy+cs
Dashboard
for
each
game
Dashboard
for
management
www.treasuredata.com
Copyright
2013
31
32. Setting Up for Governance
Cross
Game
AnalyIcs
team
can
access
/
analyze
all
data
holisIcally
Game1
only
has
access
to
their
database
Game1
Game2
Game3
Game4
Game5
A
A
A
A
A
B
B
B
B
B
C
C
C
C
C
www.treasuredata.com
Copyright
2013
32
37. 37
Project
• Treasure Data のデータコレクタ部分は、オープンソース化
• 2013年、国内外で広く浸透
• 世界中でデータを解析可能な形で収集するのに一役買っている
"We use Fluentd to collect massive data logs for our platforms. Having
developed a system based on Fluentd, we are now effectively monitoring and
analyzing our services in real-time. We are very much satisfied with its flexibility,
especially how easy it is to use in tandem with other systems."
"We utilize Fluentd to collect a very large amount of logs. The logs are
written into Hadoop HDFS clusters, and are also used to analyze
various service statuses in realtime. We also use many plugins from
rubygems.org to further enhance this mechanism."
Fluentd is very similar to Apache Flume or Facebook’s Scribe
[but] it’s easier to install and maintain and has better
documentation and support than either Flume or Scribe”
Fluentd
オープンソースプロジェクトのユーザー例