Karan Sathiayan

Experience

• 

  Perfios

  May 2019 - May 2019

  Product Engineering Intern

  In this internship, I implemented the module “Classification using Random Forest” with Machine Learning where I worked on large datasets available on ListenData and Kaggle and used an appropriate prediction model to resolve the problem statement.I also Implemented a Computer Vision task on the ‘Vahan’ Government website to simplify the captcha present in the website; used image processing tools to make the text on the Captcha image more legible and then a Google-based OCR tool - ‘PyTesseract’ to effectively read the captcha. Show less

• 

  Tata Consultancy Services

  Dec 2019 - Apr 2020

  Industrial Analyst

  Engineered a Python-based chatbot for a Retail Store Manager, providing seamless access to key sales insights and product analyticsDeployed the chatbot on AWS using Amazon Lex integrated with Lambda functions, increasing security and operational efficiency by 30%

• 

  Actyv.ai®

  Dec 2020 - Apr 2021

  Led a 4-person team, fostering cross-functional collaboration, and slashing project delivery time by 20% through roadmap formulationGenerated impactful Tableau and Power BI dashboards, accelerating decision-making by 25% Achieved 89% accuracy in Shelf object detection and recognition, leading to optimized inventory managementArchitected ETL pipelines, reducing latency by 40% and increasing operational agility for real-time object prediction

  • Senior Data Scientist

    Mar 2021 - Apr 2021

  • Trainee Data Scientist

    Dec 2020 - Feb 2021
• 

  Tata Consultancy Services

  Sept 2022 - Apr 2023

  Data Analyst

  Employed Random Forest to categorize support tickets into 20 groups with 90% accuracy, decreasing 3,000 man-hours on issue resolutionEnhanced precision for kidney image classification to 94%, leading to improved diagnostics and substantial cost savingsCollaborated with cross-functional teams to implement a streamlined pipeline for kidney classification website, boosting user experienceHarnessed Quantum Computing in Stock Market forecasting attaining a 10% faster performance than conventional methods Show less

• 

  Labelmaster

  Jan 2023 - Apr 2023

  Data Scientist Intern

  Elevated customer classification accuracy by 27% through critical thinking and strategic use of K-means and K-prototypes clusteringAutomated cluster identification, slashing quote generation turnaround time from 7 days to an impressive 3 daysRestructured pricing strategies with clustering analysis, incrementing customer quote acceptance rates by 10%

• 

  Tata Consultancy Services

  Jan 2024 - now

  Data Analyst