Kajal Singh

About Kajal Singh

I am a Data Scientist and an alumnus of IIT Kharagpur, with a strong foundation in engineering, analytics, and data driven problem solving. I specialize in transforming complex and large scale data into actionable insights that drive measurable business outcomes.In my professional role, I work extensively with Python and SQL to analyze data, build scalable pipelines, and develop robust predictive models. I have hands on experience in machine learning and deep learning, including building and optimizing neural network based solutions for real world use cases. I am comfortable working across the complete data science lifecycle, from data exploration and feature engineering to model development, validation, and deployment.I use Tableau for data visualization and storytelling, enabling clear and impactful communication of insights to both technical and business stakeholders. I strongly believe that the true value of data science lies in converting analytical results into informed and effective decisions.I actively work with and explore Generative AI and GenAI applications, including large language models and AI driven automation, to build intelligent and future ready solutions. Continuous learning and adapting to emerging technologies are core to my professional approach.I enjoy collaborating with teams to solve challenging problems and deliver scalable, data driven solutions. I am always open to meaningful conversations around data science, machine learning, deep learning, neural networks, and Generative AI. Let us connect and create impact through data.

Experience

1. Data Scientist

   SymphonyAI

   May 2025 — Present · Bengaluru, IN

   Built drilling optimization pipeline ingesting multi-rig XML logs, handling parsing, quality checks, forward-fill nulls, outlier detection on 1M rows• Engineered 15+ hole-level & 25+ bit-level features using EDA aggregation & smoothing across 100+ parameters including ROP, torque & RPM• Optimized Rate of Penetration using XGBoost, benchmarked via PyCaret & Golden Batch frameworks, achieving MAE of 0.09 & RMSE of 0.12• Validated ROP recommendations using simulator-based modeling, achieving Feed Force MAPE 0.045, Torque 0.071, and Rotation Speed 0.03• Delivered 18% reduction in average drilling time from 45.7 to 37.1 minutes, saving 8.6 minutes per hole across 80% of all evaluated holes• Designed ordinal remaining life labeling from 0 - 100% to classify drill bits into five standardized health classes for maintenance decisions• Developed five-class drill bit shelf-life model using ExtraTrees ensemble classifier, achieving F1-score up to 0.77 on extreme health states

Education

• Indian Institute of Technology, Kharagpur

  B.Tech + M.Tech