Naveen Kumar J

+918754583104 naveenharry03@gmail.com Chennai, Tamil Nadu, India github.com My Portfolio linkedin

SUMMARY
7+ years experienced, meticulous & result-oriented in senior Quality assurance and AI/Data Scientist roles armed with a proven track record of analytical
acumen in deploying complex machine learning and statistical modeling algorithms/techniques for identifying patterns and extracting valuable insights.
Possesses diverse experience in planning & executing multiple projects and liaising with the key stakeholders to identify & resolve business problem
statement and deliver excellent results. Proficient in DL Networks, Generative AI and seamlessly worked with MLOPS Culture.

KEY SKILLS
• Account Mining • Applied Research • Collaborative • Data Analysis • Data Mining & Analytics • MLOPS • Debugging
• Decision Making • Finance Domain Knowledge • Flexibility and Adaptability • Innovative Thinking • Machine Learning Methodologies
• Optimization Techniques • Patience and Perseverance • Pattern Recognition • Predictive & Statistical Modeling • Predictive Modeling & Analytics
• Problem Solving • Process Improvement • Project Delivery • Project Management • Risk Management • Sentiment Analysis • Strategic Thinking
• Team Coordination & Leadership • Team Management • Text Mining • Time Bound and Time Management • Data Insights • Deep Learning
Frameworks • Model Testing• Analyzing Data• Data Manipulation• Data Visualization Techniques• Complex Data Sets

TECHNICAL SKILLS
Project Management/Methodologies: Azure DevOps, MLOps / Agile, Kanban and CRISP-DM Methodologies
Languages/ Azure Services: Python, C#, HTML, C, C++ / Azure Data Factory, Azure Databricks, Azure Synapse Analytics, Azure Storage Data Lake
Cloud Computing / CI CD Pipelines: Azure, Serverless, Azure IAAS and PAAS, MLFlow , Azure Machine Learning Workspace/Azure Pipelines, Kubeflow
Containers and its Orchestration: Docker, Kubernetes, Azure Container Registry, Azure Kubernetes Service
Monitoring Tools: Azure Application insights, Datadog, Prometheus and Grafana.
Database / Web API Development / Devices: SQL Server, Redis/ Fast API, Flask, Swagger, Postman, streamlit/ IOS, Android, Laptop and Desktop.
Machine Learning / Deep Learning Networks: Supervised and Unsupervised Learning algorithms / Data Augmentation, LSTM.
NLP and Generative AI: Lexical, syntactic and Semantic Processing, LLM and Fin Bert / Microsoft Co-Pilot, OpenAI API, Prompt Engineering and ChatGPT.
Libraries/Packages: Pandas, NumPy, Scikit-Learn, Keras, Pyspark, Matplotlib, seaborn, NLTK, spacy, TensorFlow, OpenCV, stats models and Hugging face.
Framework/IDE/Data Visualization tool: VS Code, Jupyter Notebooks, Google Colab / Power BI, Excel, Azure Bot Service, Power Automate and PVA.

EDUCATION
Master of Science in Machine Learning and Artificial Intelligence Dec '21 - Dec '23
Liverpool John Moore's University & upGrad Liverpool, United Kingdom
CGPA - Yet to Receive

Post Graduation Certification in Machine Learning and Artificial Intelligence Dec '21 - Feb '23
IIIT Bangalore & upGrad Bangalore, IN
CGPA - 3.55 / 4

• Course Modules:
○ Introduction to Python | Python for Data Science | Data Visualization in Python | Exploratory Data Analysis (EDA) | Inferential Statistics
○ Hypothesis Testing | Linear and Logistic Regression | Naive Bayes | Model Selection (Bias and Variance Trade off) | Ridge and Lasso Regression
○ SVM | Tree Models | Boosting | Model Interpretation | Unsupervised Learning - Clustering | PCA | Hyperparameter - Grid Search | K - Fold CV
○ Perceptron | Feed Forward Neural networks | Back Propagation | ANN | CNN | RNN | LSTM | Lexical | Syntactic | Semantic | Topic Modelling
○ MLOPS Pre- requisite | MLFlow | Apache Airflow | Design Machine Learning Systems | Data Architecture | Data Strategy | CI - CD Integrations
B. E. in Electronics and Communication Engineering Sep '12 - Apr '16
SMIT, Anna University Chennai, IN
CGPA - 7.83 / 10

KEY DATA SCIENCE PROJECTS (Top 2 Mentioned here. Please find remaining 30+ projects in My Portfolio)
• Domain: Finance |Title: AI Stock Market Prediction Bot | Tech Stack: Python, Deep Learning - LSTM and Generative AI | Dec '23
○ Objective: An Individual people is struggling to make profitable money in the trading industry without any prior technical or domain experience.
○ Solution: Designed a deep Learning LSTM model with Fin Bert to predict future OHLC Values of Nifty50.
○ Key Achievement: Developed a model integrated with Chatbot and generative AI with accuracy as 99%

• Domain: Agriculture | Title: Potato Disease Detection| Tech Stack: Python | Nov '22
 ○ Objective: Farmers facing economic loss and crop waste due to various disease in potato plants.
○ Solution: Designed image classification with CNN and built a mobile app which helps to predict the plants has disease or not
○ Key Achievement: Predicted with hard classes whether the plants have disease or not with a score of 0.91

PROFESSIONAL EXPERIENCE
Senior Quality Assurance Engineer Jan '19 - Present
FE Fundinfo Chennai, IN
FE Fundinfo, a global leader in investment fund data and technology, has operations in 15+ countries & renders financial services to 6000+ financial advisors worldwide.

Speech Recognition AI Assistant Bot Development using OpenAI (Generative AI) | Phase 2 (Under Development) | Docker and AKS
• Conducted extensive research on Bot Development and integration with applications using Lang chain (LLM) models
• Led to remarkable results of 23% increase in user engagement and boosted efficiency with 30% reduction in time required for user queries
• Deployed as an Azure Bot Service using Docker and AKS ensuring high availability and self-healing auto recovery mechanisms if pod/container fails.

End to End Automated Cloud Machine Learning and Big Data Processing workflow using Azure with MLOPS Life Cycle
• Performed End to End automated Workflow using 4+ azure services like ASA, AD, ASDL, ADF for big data Processing.
• Executed End to End automated Workflow by combining output of entire big data processing workflow with Azure machine learning workspace.
• Managed 5+ years with MLOPS and 7+ years with DevOps parallelly via Agile and CRISP DM methodologies.

Optimization & Algorithm Development

• Revamped a Marketing algorithm using the concept of price elasticity of demand resulting in 11% increase in audience reach
• Deployed multiple loss minimization & optimization techniques reducing computational cost by £340 monthly in cloud operations
• Guided the development of financial products performance assessment using k-NN Algorithm resulting in decommissioning 5 low level products
• Spearheaded 2 scrum units to create a recommendation engine to suggest a model portfolio based on client risk management

Statistical Modelling & Analysis

• Developed segmentation models using K-means Clustering, revealing 6% expansion in new user segments for targeted marketing strategies
• Conceptualized and implemented a sentiment analysis tool to rate trustnet applications based on subjective customer reviews
• Deployed ML Models to Azure PAAS via Azure CI-CD pipelines which reduces 30% resource cost in azure cloud services

Key Achievements
• Deployed python to develop a customer segmentation algorithm for boosting sales leads and increasing market share by 24%
• Won the 1st price office hackathon by highlighting the benefits of Azure ML workspace, particularly suited for small product development.
• Reduced Time-to-Action (TAT) by 20% by proactively identifying and addressing model and data drift using monitoring tools.
Quality Assurance Engineer Jul '18 - Jan '19
Vega Intellisoft Pvt Ltd (outsourced to FE Fundinfo) Chennai, IN
Vega Intellisoft is a pioneer in outsourcing Partner services for IT companies for staff augmentation.

Image Classification Modelling

• Directed model development, validation, testing, and implementation of analytical products and applications
• Conceived an Image Classification model for Logo and document Identification saving 18 hrs. manual testing effort per Deployment

Dataset Management & Statistics Essentials (EDA)

• Employed Principal Component Analysis (PCA) to analyze collinearity and reduce the dimensionality of datasets
• Championed Pattern Recognition Techniques resulting in identifying top 30% most frequently used Portfolios by Fund Management groups
Software Test Engineer Jun '16 - Jun '18
Camberwell Technologies Chennai, IN
Camberwell Technologies is a service-based company providing software development and quality standards services to the clients across different industries.

Quality Standards and DevOps Methodologies

• Proficiently applied 3+ testing methodologies in diverse environments like Agile, Kanban, waterfall and Scrum as part of DevOps Process
• Key driver in DevOps adoption, realizing a 15% faster software release and elevating quality standards by 25% through continuous testing

CERTIFICATIONS
• Cricket Analytics on PowerBI | Mad About Sports| Oct '22
• ChatGPT Prompt Certification | Growth School | May '22
• Azure AI Fundamentals | Microsoft| Sept '21
• Azure Fundamentals | Microsoft | Jan '20