What is a data science program?

It is an approach that attempts to use data to extract knowledge that is beneficial to society. The program covers information science, statistics, algorithms, and other methods of handling data across a wide range of fields.

We convert your company's date and time transaction data into a format that can be operated within your company. We build data science programs in line with your business strategy and tactical plans; we build your company's unique data assets based on planning reference data through AI analysis.

STEP1) Requirement Definition Stage

Data Requirements Definition Phase

We will summarize the data handling requirements in line with the strategy based on the company's purpose, vision and financial targets. Daily transaction data will be the primary data, and we will summarize data acquisition methods, storage methods, operational members, utilization methods, and main objectives with an eye to annual targets and maximizing customer satisfaction.

STEP2) Identification of data issues

Identification of data issues

Regardless of the direction of the company, we will first examine the data needed to overcome the most pressing issues. We then proceed with data science based on the roadmap laid out in the requirements definition phase.

STEP3) Data Collection and Analysis Process

Data Collection and Analysis Process

Design and collect data using SQL or other coding for data acquisition and storage. If necessary, we will implement BI tools at this stage.

STEP4) Data Cleansing・Processing

Data cleansing and processing

The data will be presented in a variety of visual representations so that both frontline employees and management can understand the data. Delete and convert inappropriate or abnormal data before or after data collection, either manually or automatically (mainly vb, access, excel and javascript). We will perform these processes manually or automatically (mainly vb, access, excel and javascript).

STEP5) Reconciliation with analysis and preparation of report

Collation of analysis contents and preparation of report

We will collect materials to identify and report on data that will lead to problem improvement from the data already visualized. It is expected that the representation of data will "reveal the source data of benefits", "solve problems", "find new issues", and "generate new ideas".

AI Analysis & Data Science

Service Charges

Contact us

Data creation requests, etc.

First, we will organize the data lying in your company.

Contact us

Work 1


AI Predictive Analytics is a data analysis technique that uses statistical algorithms and machine learning to predict future results based on past performance. It can improve the efficiency of existing businesses and create new businesses.

■AI Predictive Analytics is a data science based on the construction of training data~analysis results with highly accurate predictions by automatic modeling. AI Predictive Analytics is a data science based on highly accurate predictions using automated modeling.

■Currently available tabular data can be numbers, strings, text, and dates, and the reason for the forecast can be verbalized.


Basic Functions

■Based on the basic functions such as binary classification (failure prediction, withdrawal prediction), multi-valued classification (complaint classification), numerical prediction (regression analysis), and time-series prediction (regression analysis), a wide range of data analysis can be performed through machine learning (learning, evaluation, prediction), including contract price prediction, targeting, production planning by shipment number prediction, purchase quantity determination by visitor number prediction, and arrival planning by product sales prediction


Prediction Method

■ Input: Tabular data in Excel, csv, tsv, etc. (ex.) Contract information: contract type (ex. Plan A) duration (ex. 5 years) number of payments (ex. 36 times) and customer information: age (ex. 33 years) gender (ex. male) address (ex. Yamagata City)

■ Prediction: Prediction models are built using machine learning (e.g.Past performance data: tabular data, customer data, sales data, and management data are used)

■Output: Prediction of results ((ex.) Probability of leaving at next renewal: 84.5%:)(Consider measures to prevent churn from leaving the system)


After Prediction

■ Discusses forecasts based on forecast results.

■Works with sales and corporate planning to understand correlations and other factors.

Information on Data Science

Engineering Blog

社会の人材不足課題に向き合う一つの方法 株式会社SHISEILABO


チャットボットの可能性 AIを用いたサービスは今では社内業務から自治体の顧客サービスの取引等に向けた様々なシーンで取り入れられています。 民間や自治体の顧客向けの取引サービスとして用いられているものとして代表的に取り上げることのできるサービスにチャットボットが挙げられます。 スマートフォンなど...

Read more

What is HTML and CSS?

This is M, our data scientist. In this issue, I will provide information related to the most basic of digital implementation: website creation. Ins...

Read more