Kshitij Zutshi

🚀 Technical Skills 🛠️ 👨🏻‍💻 Programming Languages: Python, Java, JavaScript, C#, SQL (MySQL), Shell Scripting, LLM Prompts ⚙️ Frameworks: ASP .NET Core, .NET Entity Framework, React, Flask, Django ☁️ Big Data & Cloud: AWS, S3, Sage Maker, Lambda, IAM, EC2, CloudWatch, CloudFormation, HDFS, GCP (Storage Buckets, BigQuery), FAST API, Serverless Framework, Kubernetes, Docker, Postman, Streamlit, Grafana, GitHub Actions, Argo CD, LLMs (ChatGPT, LLaMA, Claude3), YAML, git

Software Engineer Scale AI • Remote Feb 2024 - Present • Conducted comparative analysis of prompt responses among LLM Models (GPT-4, Claude3, LLaMA), identifying areas for code-based enhancements • Implemented reinforcement learning with human feedback (RLHF) to rate model responses, driving optimization for Open AI Project Feather and boosting AI accuracy and safety Philips Software Engineer - R&D DevOps Platform Philips • Cambridge, MA, USA Jan 2023 - Aug 2023 • Deployed .NET microservices for organization-wide repository DevOps metrics monitoring to CURL data from GitHub workflows to PostgreSQL cluster managed by Argo CD; metrics for 200+ active repositories visualized on Grafana and Tableau dashboards • Led AWS CloudFormation template deployment for automated EC2 instance deletion based on job duration (4/8/24 hrs.) or exceeding $2000 billing; Utilized CloudWatch alarms and Step Functions to enhance resource management, cost control, and operational efficiency by 90% • Scripted GitHub action YAML workflows to pull secrets from AWS secrets manager in 400+ Philips internal repositories Northeastern University Data Analyst - Awards & Grants Northeastern University • Boston, MA, USA Jun 2022 - Dec 2022 • Executed financial data analysis, forecasting, and modeling on Provost’s Office Grants data warehouse using IBM Cognos, SQL and Tableau thereby reducing data error rate by 20% and saving an average of 20 hours per week • Translated complex requirements into ad-hoc Tableau reports to senior leadership for enhanced data-driven decision-making; export funding & grants forecasting metrics on Northeastern University’s website Squark AI Machine Learning Engineer Squark AI • Lexington, MA, USA Sep 2021 - Jun 2022 · Engineered an Auto ML pipeline python script for automated structured dataset processing, analysis, and feature selection thereby reducing manual intervention by 80%; saving an average of 20 hours per week. · Benchmarked models using python multithreading and ML Model Optimization, increasing processing efficiency by 60% & reducing training time by 40% · Implemented model interpretability to calculate significant features using SHAP & Force SHAP on predictions generated by H20 AI WebileApps LLC Senior Software Engineer WebileApps LLC • Chennai, Tamil Nadu, India Jun 2019 - Aug 2021 • Developed digital Infrastructure for Indian e-Governance Portal’s Mortgage and estate core modules for government employees using React JSX, Redux, JavaScript(ES6+) and parsing JSON from PostgreSQL Database thereby increasing website traffic by 70% • Compiled code for legal document upload and text extraction for verification in citizen portal’s mortgage module, to achieve 40% higher fault tolerance in fraudulent document uploads; API endpoints testing done using Postman and Charles Proxy Infosys Software Engineer Infosys • Chennai, Tamil Nadu, India Sep 2017 - Jun 2019 • Led a team of 8 developers to reinforce the core architecture of Customer Account Opening (CAO) and Retail Diary (RD) modules for a Scottish banking client, ensuring heightened security & resilience in handling sensitive JSON data from API; Implemented eKYC (electronic Know-Your-Customer) feature based on special needs (disability options) of the customer using Python, ReactJS and PostgreSQL for fast and efficient data processing • Enhanced SQL queries to identify and analyze Customer Churn Rate (CCR) in users with special needs, resulting in a substantial 30% increase in customer lead generation and retention Q1 of 2020