Feb 2023 – Oct 2023 · 8 mos · Bengaluru, Karnataka, India · Remote

• Building Hexo to outstand in generative AI space.
• ➢ Reduced human involvement by 70%, retaining clients through loophole investigation and feature launches. Improved image generation tool by leveraging on prompt engineering.
• ➢ Customized Diffusion models for clients by fine-tuning on specific data.
• ➢ Enhanced existing generative AI models using deep learning frameworks like Pytorch, Transformers, Xformers, Segment_Anything.
• ➢ Added features like Segmentation, Background Removal, Inpainting, Out-painting, Photorealism, Color Control, Upscale, Adapts for an end-to-end pipeline.
• ➢ Worked with AWS services like EC2 instances to run big models on GPUs, parallel processing CPUs, worked with S3 bucket storages and used Sagemaker like services to deploy the machine learning models further.

Feb 2022 – Feb 2023 · 1 yr · India · Remote

• ➢ Built a news aggregator app prototype from scratch with 60% less human involvement.
• ➢ This MVP evolved around 4 vital features article summary in a vernacular format categorized in different genres with a audio button to listen the summary of the news article.
• ➢ Scraped Data of news articles from big publications like Times India, Times Magazine, TheWire, google News using BeautifulSoup, Scrapy.
• ➢ Managed a complete database using MySQL to fetch stored articles summary, their respective genre, translated text,respective audio as per user's choice.
• ➢ NLP based summarization using TextRank, BERT, GPT-3like algorithms; Cosine matrices; Libraries used: NLTK, SpaCy.
• ➢Used ML models: Naive Bayes, SVM, neural networks to classify articles in respective genres as per the context.
• ➢ To ensure vernacular format: collected user preferences data to support regional languages translation; Implemented language translation models; Fameworks: Snowball stemmer, Translate, tkinter, googletrans, google translate API (experimented between different libraries and APIs to select the best).
• ➢ Converted text to audio using GTTS API; deployed features/ ML models using APIs on AWS.

Aug 2021 – Jan 2022 · 5 mos · New York, United States · Remote

• Worked on two of Omdena's vital projects:
• FIRST PROJECT:
• "Improving Natural Disaster and Water Resource Management using Natural Language Processing"
• ➢ Led a 50-member team at Omdena, tackling a 2-month challenge to enhance Water Resource Management (WRM) with NLP and space data for startup partner WEO's sustainability mission.
• ➢ Developed a robust tool in 8 weeks, empowering WEO to swiftly extract critical information on floods and droughts.
• ➢ Applied advanced NLP, machine learning, and deep learning for insightful analysis.
• ➢ Defined and executed query topics (flood and drought evaluations, landcover mapping, urban climate monitoring) with NLP, NLTK, and word embeddings.
• ➢ Achieved comprehensive data gathering from diverse sources.
• ➢ Contributed significantly to an efficient data storage system, enabling streamlined data evaluation.
• ➢ Implemented metrics for statistical insights, emphasizing accuracy, precision, and recall.
• ➢ Demonstrated project impact through tangible outcomes: retrieval of a significant number of images and data points. Aligned goals with broader challenges of population growth, urbanization, and climatic uncertainties in WRM.
• SECOND PROJECT:
• "Anomaly Detection on Mars Using Deep Learning"
• ➢ Worked in a team of 50 and assessed them 70% of time.
• ➢ Detected anomalies on Mars Surface (86% confidence)
• ➢ Preprocessed Image: annotated 200 images with anomalies using the
• VGG image annotator & invented models for the project.
• ➢ Diagnosed all anomalies using YOLOv4; achieved 86% confidence
• ➢ Led 100% of the task & deployed model successfully on Stream-lit

Jan 2021 – Jun 2021 · 5 mos · Greater London, England, United Kingdom

• Built end-to-end NLP pipeline from scratch, the major work includes :
• Researched the domain and narrowed down the top potential sources to scrape data from
• Data Scraping using Python Scripts and Octoparse software
• Data Cleaning/ Normalizing /Pre-processing
• Exploratory Data Analysis
• Generating n-Grams/ Collocations
• Text clustering
• Topic Modeling
• Grammar Patterns
• Keywords Extraction for Data Labeling purpose.
• Text summarization to understand the context of the data.
• Sentiment analysis of text to build insights over the future services needed to be given.

Sep 2020 – Dec 2020 · 3 mos · Greater London, England, United Kingdom

• Worked as a Data Science Consultant to provide them with insights that will eventually help them to scale as a business.
• I have worked on Researching & Scraping Data.
• Built AI driven solution for marketing analytics in order to do better consultation.

Sep 2020 – Dec 2020 · 3 mos · Tokyo, Japan

• Octoparse is the SaaS platform that provides Scraping services.
• I have written articles around web-scraping for them.
• I have scraped all types of complex website structures such as e-commerce sites (Amazon), blogging sites, edtech-sites, OTT platforms (Netflix, Hotstar), Youtube Comments and many more not so popular websites.

Jun 2020 – Mar 2023 · 2 yrs 9 mos · India

• Sharing my learnings and knowledge by creating Data Science, ML, DL, AI related videos.
• I always wanted education to be freely available to every ends of the world and what would be better than YT to host educational videos. So, Here I am not knowing everything but here to share my learnings on the way of my journey to become world's best data scientist.
• I will keep posting every now and then about data science field, the fundamentals, the advance techs, the job roles, the kinda work we do, the new advancements or research happening worldwide and so on.
• Check out my channel- https://www.youtube.com/channel/UCdwAaZMWiRmvIBIT96ApVjw

Jun 2020 – Mar 2023 · 2 yrs 9 mos · San Francisco, California, United States

• To share my data science learnings and knowledge, I started contributing my work related to Deep learning, Machine learning, Artificial intelligence through blogs on Medium.
• So far, I have covered the basic/fundamental topics on Scraping data, Pre-processing data, Exploratory data analysis, Modelling data, Data labelling methods, Topic modelling, End-to-end machine learning projects, The stand out data science project ideas, Data Science resources, building NLP pipeline, Text clustering using K-Means, Important ML Algorithms.
• If you have any question or want to give feedback, feel free to DM me.