PRASHANT GUPTA

Generative AI Engineer:

• Internal documentation for the RAG Project: Created a knowledge base for the developers.

Tech used: web scraping, unstructured data, metadata filtering, LangChain, ChromaDB, AWS Lambda, API Gateway, Docker, and Python.

• Code generation research project: The idea was to improve generated code, test cases, and debugging accuracy, achieving approximately 30% development efficiency.

Tech used: Chain-of-Thought prompts, Langchain, Langgraph, Agentic/Self-corrective RAG, Ollma, unsloth, fine-tuning smaller code LLMs, EC2, REST API, and VS Code extension.

Data Engineering Manager:

• Lead a modern data warehouse project to capture tens of GB of daily sales and non-sales data from different data sources, and store them in various data layers.
• Reduced approximately 10% of cloud resources through stakeholder engagements, improving data models, adding materialized views, and restructuring data pipelines.
• Managing business stakeholders' expectations and guiding different technical teams.
• Follow the E2E ITSM lifecycle to act on incidents captured from automated alerts or from business users.
• Tools used: Azure Data Factory, SQL Server, Functions, Airflow, Spark, GitHub, and Python.

• Leading a data-as-a-product (DaaP) strategy project to develop data-sharing capabilities between the agencies.
• Implemented a lake house, including data cataloging, metadata, lineage, ETL, profiling, quality, classification, PII detection, governance, data modeling, data sharing, and security.
• Integrated multiple heterogeneous sources into one unified source via ETL processes.
• Monitored system health and performance metrics to ensure smooth functioning of the system.
• Implemented best practices around data governance and compliance standards such as GDPR and HIPAA.
• Tools used: Azure Purview, Data Factory/Synapse, ADLS Delta tables, Gen 2/SQL, Azure, Databricks, PySpark, DevOps, Airflow, Talend, and Python.

• Managed the data engineering and analytics team responsible for the design and implementation of a data lake architecture enriched by an ETL batch data pipeline on AWS.
• Reduced 20% of man-hours required for different manual report generation by automating data solutions for various departments, and providing live, self-served dashboards with proper data security checks.
• Collected, outlined and refined requirements, led design processes and oversaw project progress.
• Tools Used: CDC, Spark, Kafka, Airflow, Delta tables, Superset, Python, EKS, EMR, Athena, Glue, S3, PostgreSQL, Terraform, CI/CD, ERWin, and GitHub Actions.
• As part of the digital strategy, we implemented a customer data platform (CDP) for web and mobile applications.
• Expectations were to improve 20% revenue by increasing customer engagement with the trading platform by running cohort marketing campaigns and tracking user behavior through the click events.
• Tools Used: Amplitude, Figma, Product Management.
• Increased operational efficiency by enabling JIRA as a collaboration tool between the data team and product managers, including educating other department stakeholders.

• Leading agile data engineering is responsible for developing the reporting module (analytics) of a B2B broadband product used by customers from different countries to review network statistics and services offered by Singtel internationally before making the purchase.
• Prepared a new data model to store and process network data.
• Tools Used: Microservices architectures, Rest APIs, C#, Java, PySpark, CRON jobs, ETL, MongoDB, Kafka, Bitbucket, Docker, SonarQube, Prometheus, Grafana, and JIRA.

• Responsible for planning and delivering credit scoring and business analytics ML projects, helping to achieve required risk management KPIs.
• Created detailed project plans outlining milestones, deliverables, timelines, resources needed, costs estimates, and quality standards.
• Saved $100K per year licensing cost by replacing a vendor tool with an in-house built ML solution deployed on the cloud.
• Identified areas for improvement in existing processes or procedures related to software delivery.
• Established processes for tracking project progress, monitoring performance metrics, and reporting results to senior management.
• Coordinated resources across multiple teams to ensure successful completion of projects on time and within budget.
• MLOps Project: Initiated a pilot fraud detection project to identify possible credit fraud applications using ML algorithms.
• Tools used: Python, Java, scikit-learn, MLflow, Python, ML algorithms (XGBoost, etc.), REST APIs, AWS, Oracle DB, FMCA orchestration.
• Engage, influence, and communicate with cross-cultural and cross-functional teams to support end-to-end SDLC data analytics projects in the APAC region.

• Developed a roadmap and implemented an end-to-end automated trading platform that can execute synthetic spread strategies, perform risk analysis, backtesting, and strategy optimization, including real-time continuous futures price data storage.
• Using my trading and coding knowledge, I also coded, backtested, and deployed trading strategies.
• Developed tools in Python to perform statistical research, pattern matching, ML, and risk management for all proprietary synthetic fixed-income futures and equity strategies.
• ML research project: Finding patterns in time series pricing data using ML algorithms and reinforcement learning.
• Tools used: Python, PyTorch, C#, SQL, VBA, QuantHouse, CQG, Bloomberg API's, R, Power BI, TradingView, Trading Technologies, TikData.

• Responsible for developing a multithreaded C# module for trading and end-to-end trade reconciliation, which facilitates traders in booking exotic OTC derivatives trades automatically with less effort and time.
• Developed a price tool to transform data and autopopulate more than 100 fields in Excel, later used to update trades in the legacy system. Received appreciation.
• Performed troubleshooting activities for escalated support requests.
• Coordinated testing and validation procedures through software development lifecycle.
• Coordinated with business analysts from the UK and the US middle office, and the GMAG (Quant) team to clarify trade facts and cash flow validation, ensuring company standards.
• Manage stakeholders (Quant team, business analyst, testing), L2 support, UAT and production release.