Description
Fraud management Data Scientist:
The Information Technology team develops and maintains Bell’s internal systems and applications while also developing integrated technology solutions for customers across all lines of business.
The team manages Bell’s IT Infrastructure and the service availability of over 700 applications, including mission critical billing and service systems. The IT delivery teams drive the IT Work Program with over 60% of IT capital invested in initiatives that support business unit priorities and Bell’s strategic imperatives.
Bell Corporate IT is currently seeking a candidate for the position of Data Scientist to develop predictive/ detective modeling technologies and software based on integrated data from a variety of different sources. The key accountabilities for this position leveraging data science to produce fraud indicators to bolster proactive measures. The selected candidate will implement data models, perform statistical analysis, and write code to perform data exploration and deploy predictive analysis models on heterogeneous data sources in order to produce Fraud management proactive defense measures for Bell’s Fraud control system.
Position Description
A contract position in which the contractor works with both Data Engineer and Fraud management experts within Bell Canada potentially in collaboration with a Canadian university.
The Data Scientist is supposed:
To generate/ select the most significant attributes (statistical, graph and signature based) for the subscription fraud classification especially for detecting major types of suspicious account opening or extension of an existing account. Different attribute sets and various ratios of variety of traffic need to be experimentally studied in order to propose an efficient data modeling approach. To design and Implement one or more ML models to detect suspicious activities (from fraudsters) and proactively prevent start services and/ or shipping/ handing handsets in high-risk accounts. The project will involve identifying relevant models for Fraud detection and generating and selecting the most significant attributes (statistical, graph and signature based) from a variety of sources. To empirically evaluate the performance of the proposed features set and classification approach applied to both publicly available data and real Bell data. To integrating the developed algorithms into the Bell Fraud infrastructure. The successful candidate will have the opportunity to augment his/ her experience in Machine Learning, Fraud Analytics related concepts, within the scope of Canada’s largest telecommunication network infrastructure. Contractor will also have the opportunity to practice visualisation, project management and other critical business skills relevant to fraud management careers in telco industry. Key Experimental requirements include (at least 2 years):
Data Science
Apply statistical analysis and machine learning techniques to produce Fraud indicators Leverage analytics involving large datasets to evaluate, refine and improve data models and determine confidence levels for newly produced Fraud indicators Analyze subscription/ account information in order to define the current fraud landscape and further the subscription fraud risk management strategy Software Engineering Tune application and query performance using profiling tools Use scripting languages to deploy models in a production environment (e.g., SQL, Cron and Shell scripts ) Increase analytical effectiveness by implementing visualizations of fraud indicatorsProduce code to clean the data and ensure the quality of data. Knowledge of IT infrastructure and system architecture, and accessing and designing APIs or databases in order to retrieve data. Design a process for transferring this data into our existing databasesSoftware Development experience with Unix shell scripting. Experience using scripting languages to deploy models in a development environment (e.g., Cron and Shell scripts) Programming with languages such as Java, C, Python and R SQL scripting languages Working with databases including the ability to write SQL queries. Desirable experience:
Create data pipelines to efficiently run using python, scala, spark, in-memory databases Overall details: Architect and implement real-time pipelines to efficiently extract features on a scale Create unit, integration, functional tests Create performance/scalability test Deploy Statistical analytics and machine-learning predictive models Create configurable code bases and smooth deployment plans to move applications across environments Required qualifications:
Master’s or Advanced degree in Mathematics, Physics, Engineering or Computer Science or the equivalent. Experience with Data Extraction, Exploration, Transformation, Data Cleansing and Data Analytics. Experience with Big Data, Analytics, Statistical Models, Hadoop and RProgramming skills with Java, C, Python and scripting languages; Ability to interpret disparate sources of data and produce high quality intelligence assessments and reports Sound document writing skills Expose to Fraud management is an asset Existing Secret clearance or ability to obtain is an asset Ability to communicate in French is an asset Note:
The successful candidate must successfully go through extensive background verifications including but not limited to criminal record and reputational checksAll Security personnel are required to sign a letter of non-disclosure which prevents them from divulging sensitive information that they may be exposed to during their assignment. This policy is strictly enforced. Behaviour skills:
Initiative Sense of collaboration (teamwork) Effective Interpersonal Skills Compliance with commitments Results Orientation Verbal and written Supervision and monitoring Travel: up to 10% within Montreal, Ottawa and Toronto corridors depending on project needs
Work Address Details - 393 Rideau St., Ottawa (ON); on-site
