Open to opportunities

Mohit Pandey

@mohitpandey

Message

Experienced data scientist specializing in machine learning and AI.

India

Message

What I'm looking for

I am looking for a role that fosters innovation, encourages collaboration, and offers opportunities for professional growth in AI and machine learning.

With over eight years of applied AI experience, I have honed my skills in deep learning, particularly in natural language processing (NLP), computer vision (CV), and classical machine learning. My journey has taken me through various roles, including Senior ML Scientist and Data Scientist, where I have consistently driven product improvements and innovations.

At PocketFM, I played a pivotal role in enhancing user engagement through advanced recommendation systems, achieving a 4% improvement in average playtime and a 6% increase in new show listens. My work involved developing ETL processes, conducting A/B testing, and integrating machine learning models into production environments, impacting over 100 million installs. I am passionate about leveraging data to create meaningful user experiences and drive business success.

My experience extends to video analytics and neural machine translation, where I developed systems that predict virality scores and improve translation accuracy. I thrive in collaborative environments and am eager to contribute my expertise to innovative projects that challenge the status quo.

Experience

Work history, roles, and key accomplishments

Current

Neural Machine Translation

Current

NEWJ, Reliance/RIL

Feb 2020 - Present (5 years 4 months)

Developed an attention and transformer-based neural translation system for Hindi-English, achieving a BLEU score of 17 on the PMI dataset. Utilized pre-trained Flair and GloVe embeddings, and Word2Vec for vectorization.

Azure Transformers BERT NLP Embeddings

Sr. ML Scientist/Co-founding member, Data science team

PocketFM

Aug 2021 - Present (3 years 10 months)

Improved average playtime and active users by 4%, leading to a 6% increase in new show LDAU. Developed and deployed recommendation models for PocketFM and PocketNovels, powering various home feed modules and achieving significant user engagement.

Pyspark Python Airflow Sagemaker Redis Kafka Machine Learning Recommendation Systems ETL

Sr. Data Scientist/Co-founding member, Data science team

NEWJ, Reliance/RIL

Jan 2020 - Present (5 years 5 months)

Developed a video analytics engine, NETRA, for predicting video virality scores using Python, PyTorch, and OpenCV. Productionized the solution with a Dockerized computer vision pipeline and Kubeflow for load-balancing.

Python PyTorch OpenCV Docker Kubernetes Kubeflow Azure Streamlit Machine Learning Computer Vision Video Analytics

Data Scientist

UST Global

Jun 2019 - Present (6 years)

Built textual features based on an ML model to predict the probability of subrogation potential for insurance claim automation. Utilized Stanford CoreNLP and Hugging Face NeuralCoref for NLP-based feature generation and ambiguity resolution.

CoreNLP Python NLTK Hugging Face Machine Learning NLP Coreference Resolution

Machine Learning Engineer

Quantiphi INC

Feb 2018 - Present (7 years 4 months)

Developed automated attribute extraction systems for product normalization for e-commerce clients like Tommy Hilfiger and Macy's. Achieved 88% accuracy using a combination of language models, TF-IDF, and neural networks.

Python Keras NLTK GCP Machine Learning NLP Neural Networks

Data Science Intern

Analytixlabs

Jul 2017 - Present (7 years 11 months)

Derived relationships between factors and customer churning tendencies from a telecom dataset. Applied regression models to predict customer churn probability, achieving 76% accuracy.

Regression Modeling Data Manipulation Data Cleaning Statistical Analysis Churn Modeling

Education

Degrees, certifications, and relevant coursework

Coursera

Specialization, Deep Learning

Completed a Deep Learning Specialization comprising four comprehensive courses. This program provided in-depth knowledge of neural networks, machine learning algorithms, and their applications.

Analytixlabs

Certification, Data Science

Obtained a Data Scientist Level-1 Certification through a classroom program. The curriculum focused on foundational data science concepts, statistical analysis, and machine learning techniques.

Analytixlabs

Certification, Data Visualization and Analytics

Completed a certification in Data Visualization and Analytics. The program covered practical skills in VBA, Advanced Excel, Tableau, and MS SQL-Access for data manipulation and reporting.

Indian Institute of Carpet Technology

Bachelor of Technology, Technology

Earned a Bachelor of Technology degree from the Indian Institute of Carpet Technology. The program provided a strong foundation in engineering principles and technological applications.