In this role, I lead team for loan collection and account management modeling together with business support. Promote advanced machine learning techniques to business stakeholders for optimization of business operation. Help to shape the business plan and build data science road map to align with it. Communicate with legal, model risk governance, external partner, business stakeholders for model governance and approval. Collaborate with engineers and operation team for model deployment. Drive model planning, govern model build and update key stakeholders of progress and outcome. Mentor and grow talent in modeling technique and business knowledge.Key accomplishments:Increased customer contact efficiency over the third party vendor(Neustar) product via initiation and management of building best time to call model using fully innovative modeling approach at customer level.Improved all stages roll rates through precisely assign collection tools by launching customer risk model. Grew profits of good customer solicitation program via introduction of customer long term healthy model.

In this role, I direct the staff for auto pricing modeling and business support. One of my main duties is to introduce novel modeling techniques to R&D and business stakeholders, while researching business applications of AI techniques. I also build talent and foster strong connections between analytical and business teams. Moreover, I build R&D researching road map to meet business need and team growth. I create project plan and allocate resources, while monitoring project process and resolving issues; update stakeholders on project status and gain feedback for course corrections. In addition, I discuss with and learn from business stakeholders to discover opportunities for improvements.A few of key accomplishments are mentioned below:• Optimized all states profitability by $65M/year (5%) via initiation and management of building next generation vehicle symbol factors using scraped information from internet and via natural language processing and image recognition techniques.• Grew 20% profits ($50M/year) of organic family segment premium via introduction of generation concept to rating structure.• Invented modern treatment of prior incident experience, that improved customers retention with incident by over 20%.• Launched cross-department forum for all analytical teams to promote innovations, collaborations, and efficiency.• Created and spearheaded 50% of telematics research initiatives in 2020 by building cross-department team for telematics research in identifying routing and driving patterns.

During my stint here, I steered property pricing modeling, strategic non pricing modeling, and international modeling in alliance with multi-faceted business units. I also promoted advanced analytics across company and researched leading-edge technologies. Furthermore, I upskilled the teams and personnel as well as strengthened relationship with business partners. I served as a member of COE (Center Of Excellent) in Traveler. In addition to the above, I chaired routine meeting to discuss new techniques tested in different departments and shared feedback / need of new techniques from department.A few of my key accomplishments are mentioned below:• Won Travelers modeling competition of Workers' Compensation (WC) fraud detection.• Augmented conversion rate of international business by 25% through enhancements in territory price differentiation.• Improved data integrity of commercial business and slashed data error rate by 20% through anomaly detection approaches.• Evolved life time value strategy for commercial line small business to consolidate profitability and retention.• Amplified usage of machine learning techniques to 150% by driving deep research team to explore non-traditional modeling techniques, including boosting tree and deep learning; tested modeling performance under big data environment.• Boosted mid-market property pricing model segmentation power to $140M premium lift/year (10%) by inventing latest approach to utilize machine learning technique for pricing model.• Optimized commercial auto business retention by 10% via modeling on interactions between agency, customers, pricing and marketing condition.

In this role, I drove predictive modeling group in R&D and devised modeling solutions for product pricing, marketing, and retention of multi-line multi-channel commercial insurance business. I inculcated outstanding talents within team and trained product group in multivariate analysis. I was accountable for leading the efforts to build rate order calculation of Restaurant and Commercial Auto businesses from scratch.A few of my key accomplishments are mentioned below:• Grew profitability of all lines of business to $36M/year (+5%) by creating first policy level multivariate non-renewal strategy.• Enhanced pricing segmentation of business lines to $60M premium lift/year by directing renewal pricing modeling projects.• Improved customer experience and curtailed 50% of complaint rate by devising new model update methodology.• Boosted accuracy of large loss prediction to 200% by creating new parameter fitting methodology.• Invented new model version migration methodology, that improved customer experience by 50%.• Launched and oversaw modeling competition in Farmers in coordination with 13 groups across company to promote best analytical practices.

During my time here, I supported the team members to enhance analytical and modeling skills. I internally consulted for commercial insurance business, while using predictive modeling for product pricing and auditing. Moreover, I mentored junior member for work and tech development, while devising best practice for team. I communicated modeling result to business partners and assist with accurate analysis of technical results in business term.A few of key accomplishments are mentioned below:• Optimized 300% of returned premium ($4M/year) by designing commercial auto survey model.• Initiated and drove build of SAS coding library, that significantly amplified coding efficiency of team members.• Outperformed commercial modeling vendor's model by 30% ($3M/year) via proactive design and creation of workers compensation premium auditing model.• Improved 15% of commercial multi-peril new business segmentation equating to $13M premium lift/year by steering development and implementation of pricing models.