Naveen Kumaar

· Gathered complete requirements from external stakeholders to understand requirements and accordingly aligned them with the platform use cases.

· Lead the development team in defining and prioritizing critical features necessary for the system's effectiveness, translating requirements into precise technical specifications, and guiding their implementation.

· Oversee the entire development lifecycle, ensuring the timely delivery of high-quality features, and collecting user feedback to iterate on the system, enhancing its usability based on valuable insights.

· Building visualization dashboard for the various clients using Apache Superset where reporting and storytelling enabled end to end solution.

· Utilized Git for version control, managing codebase. Implemented feature branching strategy to streamline the development process. Conducted code reviews and managed pull requests using GitHub, ensuring code quality and adherence to coding standards. Resolved merger conflicts and performed regular repository maintenance.

Project Name : Valmont

• Requirement from the Valmont, in building the NER models for the Text Extraction, Modelling for the Product Taxonomy catalogue, based the business requirement and understanding the cross functionality of the domain expertise to extract responsible insight.
• Based on the Generative Open AI models and LLM models solved the issue by replacing the missing contents generated through the PERT and later extracted the Keywords and Feature for each part specific to the Product and Sub Product type.

Other Clients,POC & Hackathon :

Preventive Maintenance-HEIC,RFID False prediction,Return Prediction & Jal Jeevan.

HEIC : Perform thorough data analysis to identify patterns and trends that predict equipment failures and maintenance needs. Develop and implement predictive models using machine learning algorithms and statistical techniques to forecast performance.Implemented a CI/CD pipeline for continuous model updates and deployment.

RFID: False positives in RFID (Radio-Frequency Identification) systems occur when an RFID reader detects a tag that is not actually present in the intended reading area or when it misidentifies a tag. This can lead to inaccuracies in applications such as inventory management, asset tracking, and access control. Addressing false positives is crucial for maintaining the reliability and accuracy of RFID systems.

Return Prediction: To predict the likelihood of product returns based on various factors such as customer behavior, product characteristics, and transaction details, enabling proactive measures to reduce return rates and enhance customer satisfaction.

Jal Jeevan: Primary focus to make sure demand and quality of the water supply is achieved.

• Fraud detection and analyzing risk tax payers: By using the Python ML Algorithms and Data wrangling process in registration application detect fraud in GST and Ineligible ITC claim, providing refund only for the eligible GST members under regulations.
• Experience in algorithms Regression and classification Techniques, Naive Bayes, Clustering Algorithms, Time series data, Apriori Technique, Isolation Forest ,handling unbalanced datasets , Deep learning LSTM Techniques using Python Scripts. Later analyzing risk reward ratio in Prediction for credit card holders. Challenge was to identify customers details synchronize with outstanding balance to find the anomalies.

• Expertise in writing complex SQL Scripts in Database where loaded data from source to target by applying transformation rules , then develop, analyze stored,retrieve, preprocess data using Python.
• Developed customer interactive Dashboard using the KPI's that could produce Business insights about Overall Sales ,Key revenue,Control Exceptions,Performance metrics to respective levels and roles