Hello, I'm
Aspiring Data Scientist · Data Engineer · Analyst
Translating complex business questions into production data pipelines, predictive models, and intuitive dashboards. Specializing in SQL analytics, stochastic modeling, data integrity, and root-cause telemetry.
Computer Science undergraduate at Vellore Institute of Technology, Andhra Pradesh with engineering expertise across Data Science, Data Engineering, and Business Analytics. I bridge the gap between complex business requirements and IT solutions—architecting robust SQL extraction pipelines, validating data integrity, and engineering predictive models for strategic decision-making.
Currently working as a Project Fellow in Data Science, conducting applied AI/ML research. Analyzing multi-source, messy datasets, developing automated ETL pipelines, designing statistical experiments across ML models, and translating quantitative findings into plain-language executive reports and co-authored research manuscripts.
Oracle Cloud Infrastructure (OCI) Certified Data Science Professional with 1,000+ problems solved on LeetCode (970+ day streak). Detail-obsessed mindset: viewing unexpected data anomalies as investigative priorities to trace discrepancies to their root cause, whether tracking conversational AI call funnels or optimizing operational supply chains.
Remote / Hybrid
Analyzed 2.0M+ spatio-temporal telemetry records to forecast traffic flow and detect operational bottlenecks. Built automated SQL extraction scripts, data integrity validation, and a hybrid predictive model optimizing emergency route dispatch times by 22%.
Engineered an NLP document processing application analyzing unstructured text and conversation transcripts. Implemented entity extraction, thematic summarization, and divergence detection between agent actions and dialogue logs. Live on Vercel.
Built an end-to-end credit risk assessment model on complex financial datasets. Engineered debt-to-income and behavioral indicators, addressed class imbalance with SMOTE, and optimized Precision-Recall / ROC-AUC trade-offs to minimize default exposure.
Implemented neural network architectures for multivariate time-series forecasting. Processed multi-node sequential sensor data to model non-linear trend variance and generate high-confidence predictive scenario projections.
Designed a normalized relational database (SQL) and analytics backend for retail and supply chain order processing. Implemented automated triggers, stock level threshold monitoring, payment audit trails, and inventory turnover reporting.
Developed a multi-class predictive decision engine evaluating environmental, soil, and climate parameters. Compared statistical algorithms (Random Forest, XGBoost) to generate prescriptive crop recommendations for yield optimization.
Web analytics application tracking daily activity metrics, workout logging consistency, and user behavioral streaks. Built custom data visualization charts and progress metrics to analyze retention patterns and performance trajectories.
Developed an IoT assistive system integrating optical character recognition (OCR) with text-to-speech cognitive synthesis. Designed headless data capture, low-latency audio pipelines, and edge cognitive automation on Raspberry Pi.
Interested in discussing analytics consulting, data engineering pipelines, or data science opportunities? Feel free to reach out.
rounak.ach@gmail.com