Farshid Hossein Nejad Email & Phone Number

Iran

Singapore, Sg

Singapore, Sg

Established Automation Pipelines utilizing EMR, AWS Glue, SageMaker, PySpark, Airflow, MWAA, and Jenkins for managing nine distinct models.Designed a Demographic Model to enhance data accuracy by efficiently imputing missing values in customer telco data for nine different countries, resulting in improved decision-making and operational efficiency.Developed a predictive model for telco customer churn, which led to a significant reduction of churn rate within a six-month timeframe, greatly improving customer retention and revenue.Successfully implemented the Credit Score Model by merging Telco and Credit Bank data, achieving an impressive 0.6 Gini score using Boosting models during testing.

• Designed a Newspaper Segmentation model using PyTorch, resulting in a substantial improvement in content categorization efficiency.• Engineered a RESTful API and successfully deployed it on a local server, ensuring seamless access and functionality for Newspaper Segmentation.

Kuala Lumpur, Federal Territory Of Kuala Lumpur, My

Developed an Intelligent OCR platform that utilized Tesseract, OpenCV, Faster RCNN, and NLP algorithms, resulting in an impressive average OCR accuracy rate across diverse insurance, logistics, and banking documents. This initiative significantly reduced paperwork time and greatly enhanced operational efficiency for multiple companies in Malaysia.Designed a Fully Automated Decision-Making System that incorporated 12 Classifier Models and a visualized decision rules tree. This innovation led to a substantial increase in decision accuracy and a notable reduction in decision-making time.Implemented an efficient Model Lifecycle Management System using MLFLOW, streamlining the entire ML lifecycle. This endeavor resulted in a remarkable reduction in product runtime and delivered substantial cost savings for the organization.

Petaling Jaya, Selangor, My

Played a key role in the design and evaluation of the "Vehicle Track" system, with a primary focus on vehicle classification and Automated Number Plate Recognition (ANPR). Provided valuable insights to enhance the performance of AI models to better align with business requirements.Conducted a comprehensive assessment and provided guidance on the essential resources, including technology and staffing, required to achieve core business objectives.

Tamil Nadu, Coimbatore, In

• Conceptualized and implemented an event-driven detection system to provide optimal booking times for end-users, resulting in a significant enhancement of beauty product sizing performance.• Developed and deployed an automated Activity Segmentation protocol for dynamic packaging and cross-selling, streamlining the grouping of similar activities.• Designed and implemented an intelligent Bot Detection workflow integrated into Internet Booking Engines used by 12 prominent airlines in the Middle East and Southeast Asia, achieving a response time of less than 2 seconds.• Created a fully automated system for attribute-driven Hotel Recommendations for 5 major airlines in Southeast Asia, resulting in a revenue increase of $150,000.• Prototyped, coded, and deployed an automated Room Mapping workflow, improving room name projection accuracy and attribute extraction precision from 64.5% to 85%.• Designed and developed a semantic-driven tag extraction system for smart Hotel Ranking and Room Tag Extraction, leading to a 1.5-fold increase in daily online purchases.• Prototyped, coded, and deployed a fully automated workflow for a Room Mapping system to accurately project room names to the similar rooms, and extraction of room attributes.• Designed and coded a semantic-driven tag extraction system for the purpose of smart Hotel Ranking and Room Tag Extraction. The total quantity of online purchases was increased by 1.5 times per day.

My

Doctor Of Philosophy (Phd), Computer Science

Master Of Computer Applications - Mca, Computer Science