| | | 20,000+ Fresh Resumes Monthly | |
|
|
| |
Click here or scroll down to respond to this candidateCandidate's Name
PHONE NUMBER AVAILABLE EMAIL AVAILABLE LINKEDIN LINK AVAILABLE EDUCATIONMasters in Computer Science, University of Massachusetts Amherst Feb Street Address - Dec 2024 Relevant Coursework: Introduction to Computer and Network Security, Advance Natural Language Process- ing, Information Retrieval, Intelligent Visual Computing, Machine Learning, Theory and practice of Software Engineering, Meth Applied Stats, Distributed Operating Systems, Advanced Algorithms Graduate Teaching Assistant, working with prof. Juan Zhai as Grader for CSStreet Address : Theory and practice of Software engineeringBachelors in Computer Science, Lakireddy Balireddy College of Engineering, affiliated to Jawaharlal Nehru Technological University Kakinada Jul Street Address - May 2022 Relevant Coursework: Information Security, Data Communications and Computer Networks, Data Mining and Data Warehousing, Artificial Intelligence, Image Processing, Software Engineering, Design Patterns, Big Data Analytics, Cloud Computing, Automata Theory and Compiler Design, Design and Analysis of Algorithms, Object Oriented Analysis and Design, Data Base Management Systems, Data Structures Secured 2nd place among 300+ participants in the coding contest organized by Tecda club of IT department. SKILLSProgramming languages: C, Python, Java, R Cloud Computing: AWS Big Data Tools: Hadoop, Spark, Kafka, Hive, Cassandra, Tableau Query Languages: SQL, GQL Operating Systems: Unix/Linux Data Structures and Algorithms Web Development: HTML, CSS, Javascript, PHP Servers: Apache, XAMPP Libraries: Pandas, Numpy, TensorFlow, matplotlib, scikit-learn, PyTorch, OpenCV Tools: Git, AutoCAD, VS Code, Google colab, Slack, Code blocks, PyCharm, IDLE, Jupyter, Anaconda, Eclipse EXPERIENCEAdvanced App Engineering Associate, Accenture Aug 2022 - Jan 2023, Bangalore, IND- Designed a POC to format 50+ attributes of data corresponding to the accounting details of the client (bank).- Processed 100TB of ORC data and converted it into Parquet format using Apache PySpark. Programmer Analyst Intern, Cognizant Mar 2022 - May 2022, Bangalore, IND- Dedicated an average of 40 hours per week in learning about four new technologies, including Python libraries for data visualization, Apache Spark for big data processing, and various AWS services like EC2 and S3. Summer Intern, Swecha Jun 2020 - Aug 2020, Remote- Developed a study tool benefiting government school students, resulting in zero-cost access to learning materials.- Used Bootstrap, HTML5, javascript, MySQL to increase students interaction with the tool about 40%. PROJECTSChain-of-thought for complex reasoning using large language model. link Feb 2024 - May 2024- Evaluated the pre-trained codellama-7B instruct model from unsloth library using PEFT with LoRA as the baseline.- Fine-tuned the base model on AQuRAT dataset imported from HuggingFace containing 97,467 examples over 10 epochs using a single Nvidia A100 GPU, taking 33 hours with a batch size of 10 and gradient accumulation of 4.- Observed an increase in the models accuracy from 24.01% to 26.38%, F1-Score from 18.67% to 28.58%, the ROUGE- 1 score increased from 0.1927 to 0.7296, and the ROUGE-L score from 0.2603 to 0.6650.- Fine-tuned the base model with custom weighted loss on identical hardware settings, requiring 30 hours for 5 epochs.- Observed superior performance on this method which resulted in an accuracy of 29.92%, F1-Score of 29.71%, and ROUGE scores of 0.8323 for ROUGE-1, 0.8257 for ROUGE-2, and 0.8317 for ROUGE-L, a BLEU score of 0.648. Rocky Montana: Analyzing Violations in the State of Montana. link Feb 2024 - May 2024- Analyzed traffic stops in Montana by performing statistical tests like Chi-Square, linear regression, ANOVA to unover meaningful insights on 9 years of data using R Studio. Smart Radiology: Context-Enhanced Report Generation Using BLIP2 T5. link May 2024 - May 2024- Fine-tuned BLIP T5 model for 10 hours with lr=5e-5, a open-source model from the LAVIS library on the hon- grui/mimic chest xray v 1 dataset containing 89,400 instances which is imported from HuggingFaces dataset library.- Utilized prompt-based tuning by providing a prompt alongside the text attribute containing medical terms corre- sponding to the X-ray images to enhance the models ability to generate medical reports for non-experts accurately.- Observed Fine-tuning yields an increase in performance of approx. 300% across all metrics compared to the baseline.- Deployed the model for inference using Flask API. Personalized news categorization using large language model. Sep 2023 - Dec 2023- Refined existing LaMP models with personalization through retrieval augmentation to get accuracy of 58.9% Performance evaluation of various Machine Learning Algorithms. Feb 2023 - May 2023- Implemented and fine-tuned various machine learning models like Random Forest, Naive Bayes, Neural Networks, etc, with rigorous experiments for optimizing hyperparameters to increase accuracy on an average of 6%. Elevation based Navigation Aviation (EleNA). Feb 2023 - May 2023- Developed a navigation system adhering to MVC architecture and software design principles, enabled route opti- mization for elevation gain while limiting total distance traveled to a specified x% of the shortest path.- Implemented 10+ APIs and tested the code using JUnit which resulted in 90%+ code coverage. Distributed Systems Development: From Concurrency to Cloud Deployment. Feb 2023 - May 2023- Designed a distributed client-server systems with concurrent networking, employing both socket communication and gRPC for remote procedure calls.- Implemented REST API servers, effectively containerized micro-services using Docker, and tested distributed sys- tems for reliability, caching, replication, fault tolerance, high availability, and cloud deployment.- Deployed the microservice in AWS EC2 instance of type m5a.large to increase the latency to 400milliseconds. BERT for language processing to detect trolls. Feb 2022 - May 2022- Employed BERT to learn contextual relations of the text which resulted in 7.2% increase of detection in trolls. Face Recognition and Identification. Feb 2021 - May 2021- Implemented Haar Cascade classifier to detect facial features to speeding up the computation by 5 times. Code Review System. Jul 2019 - Nov 2019- Developed a user interface using HTML, CSS, Javascript, Bootstrap with user authentication and authorization.- Designed a SQL database schema for efficient data storage and retrieval allowing users to LEADERSHIP Managed daily activities and assemblies as School People Leader and led 500+ students. Led the schools Rotary Club as President in 2014-15, actively coordinating social welfare activities and events. Tackled challenges affecting women within the college community as the Class Representative of Saheli club. |
