Sanjay J. - NLP Engineer

Go to Website

Düsseldorf, Germany

Experience

Feb 2024 - Sep 2024

8 months

NLP Engineer

German IT consulting company

Developed an ethical, highly scalable Selenium-based web crawler that revolutionized data collection from JavaScript-based websites and significantly expanded its reach.
Designed and implemented an end-to-end pipeline for B2B lead qualification using AWS services (Step Functions, Lambda, Fargate, DynamoDB, S3) and Pinecone Vector DB to automate and optimize the process.
Built an ML-based preprocessor to efficiently filter large datasets, which drastically cut embedding and storage costs and improved retrieval accuracy and the performance of LLMs.

Apr 2023 - Dec 2023

9 months

AI Engineer

German FinTech company

Migrated a BiLSTM-CRF model to a pretrained BERT model, improving data anonymization by 12%. This led to a significant reduction in processing errors.
Developed a Natural Language Processing (NLP) model to extract loan contract attributes such as interest rates and repayment schedules, boosting contract analysis efficiency by 30%.
Introduced FastAPI to optimize API latency and enable seamless access to multiple ML services, improving response times by 40%.

Jun 2022 - Oct 2023

1 year 5 months

NLP Engineer

German mid-sized education company

Developed a high-precision Named Entity Recognition (NER) model using transformer architectures to detect learning content from module descriptions. The model achieved an 85% F1 score and was integrated into a scalable data pipeline.
Implemented an entity linking model with dense retrieval and graph-based approximate nearest neighbor algorithms, improving information accuracy by 90%.
Used Scrapy and Selenium for automated data extraction and processing to create a scalable data pipeline that increased data processing efficiency by 50%.

Feb 2021 - Mar 2023

2 years 2 months

Data Scientist / Automation Engineer

Telecommunications company

Optimized the profile analysis process for data analysts by 60% by developing a Python-based automation solution.
Accelerated the ETL code review and merge process for data engineers by 90% by integrating Oracle APEX, PL/SQL, JavaScript, and Bash.
Implemented a transformer-based email classification model to improve customer communication and reduce manual sorting time by 75%.
Automated customer updates with an IR system using OCR and NER, leading to an 80% reduction in manual work.
Led the digital transformation of a sales department and the company.

Oct 2020 - Mar 2021

6 months

AI Engineer

German research institute

Studied the impact of Gaussian noise on the accuracy of orientation detection models using Convolutional Neural Networks (CNNs) and YOLO. This resulted in a 20% improvement in model accuracy.
Developed a Deep Q-Network (DQN) to analyze learning stability and accuracy of reinforcement learning models, increasing learning algorithm efficiency by 25%.
Used LabVIEW to control and analyze experimental data in real time, reducing data processing time by 30%.

Summary

Sanjay Jayaprakash is an experienced Data Scientist and AI specialist with over 5 years of expertise in developing and implementing innovative machine learning models and data-driven solutions for various industries.

His core skills include Natural Language Processing, data integration, predictive analytics, and leading digital transformation projects.