Front-End Development: Build responsive and intuitive user interfaces using HTML, CSS, and JavaScript frameworks, with a strong emphasis on React.Back-End Development: Develop server-side logic, databases, and APIs using technologies like Node.js, Express, Python, or Java.Database Management: Design and manage relational and non-relational databases (e.g., MySQL, MongoDB) to ensure efficient data storage and retrieval.Integration: Integrate third-party services and APIs to enhance application functionality.Testing: Implement unit and integration tests to ensure code quality and reliability.Deployment: Deploy applications to cloud platforms (e.g., AWS, Azure) and manage continuous integration/continuous deployment (CI/CD) pipelines.Maintenance: Monitor and maintain applications to ensure optimal performance and security.Collaboration: Work collaboratively with cross-functional teams to understand project requirements and deliver solutions that meet client needs.

CreaDetect is a test strip and a portable electronic meter which measures creatinine from the blood drop of kidney disease patients. My job role includes data processing, feature extraction, feature selection, Ml model training for failures calculations regarding the particular domain of renal analysis. The dataset consists of images taken under the microscope along with the tabulated observations.

Problem: Rollout of 5G presented a challenge, Network slicing, the ability to carve virtual networks for specific use cases, demanded efficient resource allocation and dynamic optimization.Project Highlights: Designed a federated learning framework to train network slicing models on data distributed across edge nodes leading to efficient catering of diverse user demands on 5G network.Developed a reinforcement learning agent that dynamically allocated resources and network slices for dynamic resource management and anticipating network bottlenecks.Wrote algorithms for Network Capacity Planning, Real Time dynamic network optimization and fault detection protection. Identified Hardware, Data Pipeline, AI/ML frameworks and libraries, Cloud infrastructure and storage, Data/Model interoperability, Data Management and monitoring to support ML algorithms. Deep understanding of network infrastructures, traffic patterns and performance metrics. Developed an LSTM based model to predict future network traffic patterns based on daily real time data, user behavior and external factors like weather and events. Achieved 36% improvement in capacity planning accuracy compared to traditional methods. Designed a learning system to analyze real-time network traffic data and dynamically adjust network parameters resulted in 21% reduction in operational costs optimizing resource utilization and preventing infrastructure upgrades. Proposed anomaly detection algorithms to identify potential network issues and failures in real time enabled faster problem identification and resolution minimizing downtime by 8%.Reviewing and implementing changes to existing infrastructure to enhance reliability, increase capability and ensure maximum availability of network infrastructure.Monitoring network stability by modifying network infrastructure in response to application changes.How to install, configure and maintain network services, equipment and devices.

* Predictive Maintenance with Edge AI and Sensor Fusion:Develop an embedded system prototype to process various sensor information on drilling equipments and pipelines. Collect and pre-process historical data incorporating failure records. Training in-house neural network models optimized for edge devices to predict equipment failures and recommend maintenance actions. Designed data pipeline to transfer sensor data from edge to cloud for data analysis and anomaly detection increasing proactive maintenance, optimized maintenance schedules and lowering operational service costs.* Reservoir Characterization with Deep Learning and Seismic Interpretation:Train a convolutional neural network to analyze 3D seismic data for more accurate reservoir characterization, identifying potential drilling locations with higher yield. Integrated model’s predictions with existing geological models for reduced exploration costs by avoiding unproductive wells. * Real-Time Production Optimization with Reinforcement Learning: Implemented a reinforcement learning algorithm to optimize production parameters in real time maximum output while preventing equipment damage. Training a reinforcement learning model to learn optimal control strategies through simulation and reward feedback. Developed a real-time control system that interacts with wellbore equipments. Train an autoencoder and models to learn normal operational patterns, identify potential deviation anomalies and trigger event alerts.

• Spearheaded the development of customized drilling of oil reservoirs with 2D/3D geostatistical simulation and soil feature classification achieved an outstanding 95 percent accuracy and yield 83 percent better results with remote sensing systems.• Engineered data pipeline to acquire raw data from remote sensing devices, cleaning and preprocessing irrelevant noise outliers, custom feature extraction algorithms with geospatial data processing, model development, scalable distributed Cloud computing infrastructure setup, reliable data storage systems, derived actionable insights to support decision making from visualizations.• Enhanced 57 percent operational efficiency, 24 percent reduced exploration costs, and efficient integration with teams.