Ayush Raj

Detail-oriented Information Technology B.Tech student with expertise in data analytics, process mining, and Android OS development. Proven ability to optimize data workflows and implement ETL processes, as demonstrated by migrating analytical queries that resulted in daily cost savings of approximately $1,000 at Slice. Proficient in Python, SQL, data analysis, cloud technologies (AWS) and Automation (Airflow).

• During a critical three-way merger, migrated analytical queries from MSSQL (CBS LAYER) to Spark.SQL, consolidating data workflows and generating approximately $1,000 in daily cost savings ($365K annually).
• Designed and developed complex SQL scripts totaling over 10,000 lines of code for Anti-Money Laundering (AML) rules, creating a new table schema to store calibrated threshold values and generating alerts for suspicious activities.
• Developed advanced SQL DAGs in Apache Airflow to orchestrate complex ETL processes, monitor daily metrics, and generate reports.
• Developed an innovative tracking system that pulled IP geolocation data from four relational databases to monitor Slice app login/logout activities across devices, enhancing security monitoring capabilities.
• Debugged and refactored numerous legacy code segments to ensure data integrity and consistency during the migration process.
• Leveraged AWS services (EMR Studio, S3) and PySpark to build cloud-native data solutions aligned with the company's digital transformation strategy.
• Collaborated cross-functionally with integration teams to troubleshoot data discrepancies and ensure analytical accuracy throughout the transition.

• Data Structures
• Computer Networks
• Object-Oriented Programming
• Database Management Systems
• Operating Systems
• Cloud Computing
• Parallel and Distributed Computing

• Process Mining for Business Process Improvement:Converted raw SQL data dumps into event logs to facilitate business process analysis and optimization
  Applied Object-Centric Process Mining techniques using PM4Py to uncover patterns and enhance process workflows
  Published research paper, From Chaos to Clarity: An Object-Centric Process Mining Case Study in Complaint Requisition Management Systems, in CEUR Workshop Proceedings 2024
  Technologies used: Python, PM4Py, NumPy, Pandas
  
• Docbot: Built a medical chatbot using Retrieval-Augmented Generation (RAG) with quantized Llama-2 LLM and vector database knowledge base. Technologies: Python, Langchain, Hugging Face, Flask, Pinecone VectorDB.
• Microstack Container Deployment: Deployed and managed web services on Ubuntu Microstack containers using OpenStack for cloud resource optimization
• Credit Card Fraud Detection: Created real-time fraud detection system processing streaming transaction data. Technologies: Apache Spark, Kafka, PySpark, Python.

Research Paper" From Chaos to Clarity: An Object-Centric Process Mining Case Study in Complaint Requisition Management System" - CEUR Workshop Proceedings (2024):

https://ceur-ws.org/Vol-3758/paper-31.pdf

• Public Relation Head in organizing Team of Technical Fest APAROKSHA IIITA.