Summary
Overview
Work History
Education
Skills
Interests
Work Preference
Timeline
Hi, I’m

Vishal Lakha

Machine Learning Engineer
Bengaluru,Karnataka

Summary

7 years of work experience in Data Science | IIT Kharagpur Graduate | Expert in delivering end-to-end Scalable Machine Learning models | Experience in NLP, Computer Vision, AWS, Databricks, SQL

Overview

8
years of professional experience

Work History

AngelOne
Bengaluru

Machine Learning Engineer
02.2022 - Current

Job overview

  • Worked effectively in fast-paced environments.
  • Self-motivated, with a strong sense of personal responsibility.

Hewlett Packard Enterprise Bengaluru

Data Scientist (Specialist)
01.2019 - 01.2022

Job overview

Robert Bosch Engineering and Business Solutions Bengaluru

Data Scientist
01.2017 - 01.2019

Job overview

TATA Steel Jamshedpur

Intern
01.2016

Job overview

  • Projects
  • Computer Vision
  • Developed a PAN Card OCR application with Paddle OCR and RetinaFace, achieving a latency reduction of 1 second compared to the standard vendor solution
  • Implemented RGB color proportion and a decision tree algorithm to create a robust signature detection model, achieving 89% F1 Score
  • This initiative led to 20% reduction in manual verification costs
  • Trained Tiny-Yolo model tailored for detecting Indian-style license plates, achieving 86% recall rate and 0.95 Inter- section over Union
  • Successfully deployed the solution on edge devices for efficient real-time performance
  • Utilized a combination of Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Connec- tionist Temporal Classification (CTC) techniques to develop a classifier for recognizing characters on license plates
  • Natural Language Processing (NLP)
  • Built a custom Named Entity Recognition (NER) Model utilizing BERT embeddings, achieving 95% recall rate for
  • Advisory Documents
  • Implemented a pre-trained Siamese Neural Network model from HuggingFace to assess document similarity against server specifications, employing cosine similarity metrics
  • Achieved 92% accuracy in generating a tailored selection of relevant advisory documents for clientele
  • Conducted data collection by leveraging Selenium for web scraping and stored the gathered data in MongoDB
  • Personalization
  • Implemented Epsilon Greedy Multi-Arm bandit & Lin UCB algorithm to dynamically select and personalize widgets on the homepage based on individual user activity within the app
  • Monitored and analyzed user interactions to fine-tune the bandit models, resulting in improved user satisfaction and lift in click-through rates using A/B Testing
  • Processing high volume data in scale of 15 crore+ daily data points of mobile application user activity using PySpark
  • Deployment and MLOps
  • Developed Flask API endpoints for model serving, along with JWT authentication mechanism for better security
  • Created Docker Images and established automated procedures for storing them in Amazon Elastic Container Registry (ECR) using GitHub Actions
  • Orchestrated the deployment of the applications on Amazon ECS clusters using Terraform scripts
  • Also constructed various monitoring dashboard utilizing Grafana for comprehensive performance oversight
  • Papers & Academic Projects
  • Abstract Presentation: Text Mining based prediction model for incident occurrences in steel plant at INFORMS Annual
  • Meeting Conference 2016 Nashville, USA
  • Performed comparative analysis of classification task using SVM and Random
  • Forest algorithms and k fold cross validation
  • Publication: Sarkar, S., Lakha, V., Ansari, I., & Maiti, J
  • (2017)
  • Supplier selection in uncertain environment: a fuzzy
  • MCDM approach
  • In Proceedings of the First International Conference on Intelligent Computing and Communication (pp
  • 257-266)
  • Springer, Singapore.

Education

Indian Institute of Technology

Bachelor of Technology from Industrial Engineering
2017

University Overview

Skills

  • Relevant Skills And Courses
  • Languages:
  • Python, SQL, PySpark, C
  • Data Science: Machine Learning, Deep Learning, Tensorflow, Keras, CNN, LSTM , Regression, Classification
  • Tech Stack: DataBricks, AWS, Git, Docker, VS Code, Linux
  • Courses: Probability & Statistics, Data Analytics, Optimization Methods, Operations Research, Linear Algebra

Interests

Extra Activities , Kaggle Competitions Digital Painting, Graphic Designing @ behancenet/vishallakha

Work Preference

Work Type

Full Time

Work Location

HybridRemoteOn-Site

Important To Me

Work-life balanceWork from home option

Timeline

Machine Learning Engineer
AngelOne
02.2022 - Current
Data Scientist (Specialist)
Hewlett Packard Enterprise Bengaluru
01.2019 - 01.2022
Data Scientist
Robert Bosch Engineering and Business Solutions Bengaluru
01.2017 - 01.2019
Intern
TATA Steel Jamshedpur
01.2016
Indian Institute of Technology
Bachelor of Technology from Industrial Engineering

Similar Profiles

CREATE PROFILE
Vishal LakhaMachine Learning Engineer