Pranav Shridhar Kowadkar

• [RAG, Python, OpenRouter] Engineered a retrieval-augmented generation pipeline using custom chunking and embeddings, delivering 31% faster document search and 18% higher QA accuracy over vanilla LLM baselines on neuroscience datasets.
• [Neurosynth, stDNN, LangChain] Developed "brain-cog-rag" by integrating Neurosynth's neuroimaging data with structured DNN retrievers, boosting relevant fact retrieval by 23% in cognitive science use-cases.
• [PyTorch, fMRI, Transformers] Designed fMRI tokenization pipeline converting 90k+ voxel time-series to transformer-friendly embeddings, enabling sequence-based cognitive state detection.
• [Attention, Explainability, NLP] Built attention mechanisms to spotlight clinically significant brain regions, raising interpretability scores by 28% in expert reviews.

• [Oracle, PL/SQL, Unix] Revamped Oracle-driven data pipelines using advanced PL/SQL and Unix scripting, boosting query speed by 20% and streamlining batch financial processes.
• [Prometheus, Monitoring, Exporters] Deployed custom Prometheus exporters for Oracle, capturing KPIs like query latency and tablespace usage to drive real-time performance monitoring.
• [Python, Code Quality, Automation] Refactored core Python modules for ETL workflows, slashing code review time by 15% and raising deployment stability.
• [Grafana, Dashboards, Alerting] Built Grafana dashboards and set proactive alerts, reducing system outages by 45% through real-time visualization of pipeline health.
• [AWS Lambda, CloudWatch, S3, RDS] Engineered end-to-end observability for Lambda data flows using CloudWatch + Grafana, accelerating root-cause analysis of bottlenecks.

• [SQL Workbench, Python, Pandas, NumPy] Automated data cleaning and analysis of 1M+ records using advanced SQL queries and pandas workflows, accelerating insights and slashing manual processing.
• [Airflow, AWS EC2, Automation] Designed, deployed, and maintained Airflow pipelines on AWS EC2, automating reports and reducing manual effort by 40% while boosting operational efficiency by 25%.
• [Looker Studio, Dashboards, Data Viz] Developed executive dashboards in Looker Studio, visualizing program impact and student engagement to drive better leadership decisions.
• [Jira, Agile, Collaboration] Drove agile project management using Jira for sprint planning and cross-team sync, keeping all deliverables on track and transparent.
• [Scrum, Kanban, Process Optimization] Applied Scrum and Kanban to balance workload and break system bottlenecks, improving throughput and adaptability across teams.

• [Databricks, Spark, Python, Snowflake] Built end-to-end Databricks pipelines to process Snowflake healthcare data, producing 3 marketing personas and guaranteeing high data reliability.
• [PowerBI, DAX, Data Visualization] Developed PowerBI dashboards with advanced DAX, visualizing 121+ behavior metrics and driving a 22% ROI boost through campaign insights.
• [Azure Data Factory, ETL, Python] Optimized real-time ETL for point-of-sale data with ADF and Python, using KPIs to enhance advertising strategy and lift campaign performance by 18%.
• [Cross-Functional Collaboration] Partnered with marketing and sales teams to turn business needs into technical specs, ensuring solutions hit every organizational goal.
• [Automation, Data Quality, Compliance] Automated data quality checks across pipelines, ensuring end-to-end integrity and compliance with privacy standards.

• [C/C++, Memory Profiling, CATIA] Fixed 10+ critical memory bugs in CATIA apps using C/C++ and profilers, cutting crash rates by 20% and boosting stability for enterprise users.
• [Linux Migration, Debugging, Integration] Led CATIA migration to Linux by resolving 20+ platform errors, slashing migration downtime by 40% and ensuring seamless rollout.
• [Python, Test Automation, CADAM] Automated CADAM testing with a Python framework, shrinking execution time by 87% and increasing code coverage by 35%.
• [Python, Pandas, Matplotlib, Analytics] Automated team analytics—processing JSON data into reports, saving 3+ hours/week and driving data-backed decisions.
• [Apache, Dashboards, Self-Service Reporting] Built Apache-hosted reporting dashboards for contributor achievements, now used by 8+ departments for leadership reviews.
• [Statistical Analysis, Product Strategy] Applied statistical analysis to software usage data, surfacing trends that shaped product development priorities.

• [CFD, Python, Data Analysis] Boosted VTOL UAV design efficiency by 17% via manual CFD and Python stats, optimizing wing configs from 1,500+ wind tunnel data points.
• [Tableau, Python, Visualization] Accelerated propulsion modeling 13% by building interactive Tableau dashboards with matplotlib/seaborn for real-time efficiency analysis.
• [CATIA V5, 3D Scanning, Compliance] Improved propeller performance by 8.5%—optimized 12 blueprints using CATIA V5 and 3D scan data, meeting MIL-STD structural standards.
• [Python, Automation, Reporting] Automated stress-strain reporting by replacing Excel with Python scripts, cutting report time from 6 hours to 45 minutes at 99.8% accuracy.
• [Machine Learning, Design Prediction] Applied ML to predict design-driven performance, making iterative UAV development faster and more data-driven.

• [C#, SQL Server, Automation] Modernized legacy banking by building a C# transaction engine with SQL Server, automating exception handling to cut manual reconciliation by 75% and speed up ₹1.2M+ daily settlements by 25%.
• [C#, OOP, Multithreading] Refactored C# financial apps with OOP and multithreading, dropping batch processing time from 6 hours to 4.5 hours for 50K+ records while keeping 99.9% data integrity.
• [SQL, Optimization, Reporting] Resolved database bottlenecks—redesigned 15+ stored procedures and indexes, slicing query latency by 35% for client-facing, high-traffic reports.
• [Python, Data Analysis, Fraud Detection] Developed scripts to spot transaction patterns, surfacing insights that improved fraud detection across banking products.
• [Automated Reporting, Business Intelligence] Built automated reporting pipelines that converted raw financial data into actionable intelligence for faster, smarter business decisions.

Completed a comprehensive program focused on advanced data science techniques, machine learning algorithms, and big data technologies.

Developed a strong foundation in mechanical engineering principles while exploring interdisciplinary applications in aerodynamics and computational design.

• Implemented collaborative filtering and content-based recommendation algorithms
• Designed a real-time feature engineering pipeline using Kafka and Spark Streaming
• Optimized model serving with Redis caching to achieve 50ms average response time
• Deployed using containerized microservices architecture on AWS ECS

Engineered a system to extract insights from unstructured financial PDFs using transformer-based LLMs fine-tuned on domain-specific data.

• Utilized OCR and PDF parsing to extract textual content from scanned financial statements
• Leveraged fine-tuned BERT and GPT-4 models to summarize and answer questions on earnings reports
• Built an RAG (Retrieval-Augmented Generation) pipeline with Pinecone vector search and LangChain
• Deployed system with fastAPI and integrated GPT-based feedback loops to improve accuracy

Developed an advanced analysis pipeline for gravitational wave detection data, implementing machine learning algorithms to identify potential signals in noisy detector data.

• Designed specialized neural network architectures for detecting faint gravitational wave signals in LIGO data
• Implemented time-frequency analysis techniques to characterize signal properties
• Achieved 94% accuracy in detecting simulated gravitational wave events
• Created interactive visualizations to explore detection results and waveform characteristics

Created an interactive web-based dashboard for exploring and analyzing anime ratings, popularity trends, and genre preferences across different demographics.

• Scraped and processed data from MyAnimeList using Python, creating a dataset of 10,000+ anime titles
• Developed interactive visualizations for exploring trends in genre popularity over time
• Implemented recommendation features based on user preferences and viewing history
• Deployed the application using Streamlit Cloud for public access