Chaitra Vishweshwaraiah

• Enhanced security features and data models, resulting in a 25% decrease in security breaches and improved customer trust.
• Improved Honeywell route manager reliability by 10%, thereby minimizing unexpected downtime for customers through automated analysis and prediction of tool failures using specific parameters from the consolidated dataset.

• Improved data processing efficiency by 40% through implementing semantic search for tax documents using Generative AI, resulting in quicker information retrieval and analysis.
• Developed data pipeline utilizing Azure Document Intelligence, with 20% reduction in data processing time.

• Reduced maintenance costs by 20% through the implementation of ML algorithms for failure predictions of machines and components, leading to increased operational efficiency for solar and wind farms.
• Led team to execute web scraping solution with 95% data extraction efficiency, providing crucial Air Quality Index data for ML operations resulting in improved decision-making and forecasting.
• Saved 6 hours/week of engineering time by eliminating traditional manual processes through the development of an ML/NLP solution that automated data attribution from raw sales data to a structured database for analytics.

• Increased detection scenarios by 55% from 20 to 36 for an AML project, reducing reputational risks and enabling Fraud Detection for a finance-based client, resulting in enhanced automated AML processes.
• Increased data-informed decision-making by 30% through creation of interactive dashboards with visualizations, resulting in improved business performance.

Developed a system to optimize the end-to-end recruitment pipeline, tailored to aid recruiters and sourcers. Automated key stages, including resume screening, candidate shortlisting, and interview scheduling, all while maintaining role-based access controls to uphold data security and privacy.

• Python
• MySQL
• Amazon S3
• HTML
• CSS
• Javascript
• Okta

"Skip to a Human" streamlines the process of getting in touch with customer service representatives. Users simply input the customer service number they're trying to reach, and our app bypasses the automated IVR systems, directly connecting them to a human being.

• GPT-4
• Google-web-speech-api
• Typescript
• node.js
• express.js
• ngok
• Websockets
• Twilio

Finetune GPT-3.5-turbo to transcribe the text for the podcast, Implemented eleven lab TTS for voice generation with customizable male and female voices to generate podcast for Machine Learning with minimum human intervention.

• Python
• YOLO

Designed a model to analyze existing data and predict the price of insurance premiums for a potential customer based on their health and other personal attributes. Uses the Random Forest Regression algorithm.

• Python
• Featurewiz
• Pycaret

Designed, Developed gym website using React, Django and deployed it on AWS EC2 for seamless accessibility ensuring agile project management through Atlassian JIRA's Scrum framework.

• Front End: React (Typescript) + Bootstrap
• Back End: Django
• Database: PostgreSQL
• Deployment: Amazon Web Service, Docker
• Scrum: Atlassian JIRA

A Guide on Using Recommendation System to improve User Experience in E-commerce platform

Uncover the world of large language models! Large Language Models (LLMs) have proven to be powerful in various tasks but fine-tuning them requires extensive supervision. In contrast, humans can improve their reasoning abilities through self-thinking without external inputs. In this article, we explore how an LLM can self-improve its reasoning ability without supervised data.

Unlock the potential of prompt engineering! This article delves into the what, why, and best practices of crafting effective prompts for AI models. Explore the art of designing prompts that yield desired results and gain valuable insights into optimizing interactions with language models, making it a must-read for those seeking to harness the power of prompt engineering.

In an ever-evolving world, understanding population dynamics is pivotal. Population numbers are more than just statistics; they represent social, economic, and political narratives of nations. From shaping policies to forecasting needs, predicting population trends holds the key to anticipating challenges and preparing for the future.