Overview
Work History
Education
Skills
Websites
Projects
Timeline
Generic

SAI SISTLA

Chennai

Overview

1
1
year of professional experience

Work History

Application Support Engineer

Prodapt Solutions
Chennai
05.2022 - 09.2023
  • Resolved over 100+ end-user incidents, reducing average issue resolution time by 25% and enhancing customer satisfaction through timely technical support.
  • Monitored application and server performance daily, proactively identifying and reporting issues that reduced unplanned downtime by 15%.
  • Streamlined user access management by deprecating 50+ obsolete profiles and roles, strengthening system security and reducing access-related errors by 20%.
  • Investigated database logs to detect memory inefficiencies and navigation bottlenecks, contributing to a 10-15% improvement in application response time.
  • Led the deployment of system updates across production and staging environments, supporting over 20 successful rollouts without rollback incidents.
  • Optimized memory usage by executing stored procedure-based cleanup on critical tables, decreasing database size by 30% and improving query speeds by 40%.
  • Conducted comprehensive VPN compatibility checks across 10+ user environments, improving remote access reliability and reducing connection issues by 35%.

Machine Learning Intern

Nexus Info
  • Developed a modular, console-based chatbot application with support for multi-turn dialogue and memory recall, simulating human-like interaction and improving session continuity by over 70%.
  • Designed a conversation context memory system using Python lists to track and recall up to 10 previous user-bot exchanges, enabling dynamic and informed replies.
  • Engineered a customizable interaction model using ask user custom to blend predefined and user-specified questions, enabling the bot to adapt dynamically to diverse conversational contexts.
  • Integrated a basic NLP-like rule-based layer for handling common questions and fallback responses, achieving an accuracy of 90% in recognizing predefined questions during testing.
  • Implemented robust error-handling mechanisms through a dedicated handler function, preventing crashes and reducing misinterpreted inputs by over 80%.
  • Implemented a conversation summary feature that displays full interaction history, boosting transparency and allowing users to track queries and responses in real time.
  • Designed and tested the chatbot's core loop to support continuous input and interaction until manual exit, resulting in an average session length of 8-12 user inputs.
  • Deployed machine learning models, including Random Forest and Logistic Regression, to identify fraudulent transactions with 92% accuracy.
  • Applied algorithms like Logistic Regression, Decision Tree, Random Forest and KNN algorithms to analyze the accuracy of the model.
  • As the dataset is quite imbalanced, techniques such as SMOTE, ADASYN, RANDOM-OVERSAMPLER and RANDOM-UNDER-SAMPLER techniques to improve accuracy.
  • Conducted model comparisons using Stacking Classifier, Voting Classifier, Bagging Classifier to analyze the overall accuracy of all models used.
  • The accuracy received by Decision Tree was 0.95, for Logistic Regression it was 0.94, for KNN the accuracy was 0.82.
  • Visualized Precision-Recall and ROC curves, adjusting thresholds to balance precision and recall for both classes.
  • Technologies Used: Python, scikit-learn, Pandas, Matplotlib, Imbalanced-learn.

Education

MS - Computer Science

Seattle University
Seattle, Washington
12.2025

Bachelor of Technology - Computer Science

GITAM Deemed University
Hyderabad, India
04.2022

Skills

  • Python
  • SQL
  • Java
  • NumPy
  • Pandas
  • Matplotlib
  • Seaborn
  • Apache Spark
  • Hadoop
  • Hive

Projects

Autism Screening Detection in Adults, Analyzed the screening scores for adults on the scale of 1-10 and based on the person's whether the related person has autism detected or not., Used classification algorithms Logistic Regression, KNN and SVM algorithms with accuracies of 52% for Logistic, 89% for SVM using polynomial kernel and 80% for KNN with K=10 values., For the Linear Regression model, the best accuracy was 32%.

Timeline

Application Support Engineer

Prodapt Solutions
05.2022 - 09.2023

Machine Learning Intern

Nexus Info

MS - Computer Science

Seattle University

Bachelor of Technology - Computer Science

GITAM Deemed University

Similar Profiles

CREATE PROFILE
SAI SISTLA