Gaurangi Raul

Software engineer with ~2 years of experience building and deploying Python backend services on AWS (SQS, S3, Lambda).Built and shipped a TypeScript + React SaaS from scratch; currently extending a multi-agent MCP system in production.

AI Engineer Intern - Right Skale Inc. - Pleasanton, California

(2026-01 - 2026-05)

• Instrumented a multi-tenant AI pipeline serving 3+ enterprise clients with Langfuse to track token usage, inference cost tracking, and latency, identifying inside bottlenecks that standard application logs missed in the distributed system
• Refined prompt templates and model-routing logic based on trace analysis, reducing LLM inference costs by 40-50% and improving end to end response latency for enterprise model while preserving output quality.
• Built asynchronous processing with AWS SQS to queue document processing across S3-backed pipeline stages, decoupling ingestion from inference so doesn’t slow down the response.
• Supported migration to FastAPI and containerization with Docker on AWS; used PostgreSQL and Playwright MCP integrations by checking TCP/IP connectivity, database host and port settings, and application logs to identify connection failures and validate fixes.

Research Data Scientist - University of Arizona - Tucson, AZ

(2025-01 - 2025-07)

• Developed an end-to-end RAG pipeline spanning ingestion, chunking, and vector retrieval using LlamaIndex and PyTorch-based embeddings to support high-performance cybersecurity workflows through hybrid ranking algorithm in Python combining semantic similarity and keyword scoring, improving ranking quality by 25% over baseline retrieval.
• Built an evaluation framework to validate ranking pipeline, achieving 100% faithfulness, retrieval Hit Rate outperforming baseline GPT-3.5 variants.
• Scale machine learning utility by integrating zero-shot LLM-based sentiment analysis on VR learner interactions, improving contextual relevance and boosting learner engagement with course content by 30%.
• Contributed to cross-functional system design with stakeholders and instructional designers to align platform output with production performance standards.

Software Engineer - Internshala - Remote · Mumbai

(2023-10 - 2024-04)

• Designed and built SQL data pipelines for ingestion and A/B testing, collaborating within an Agile framework (sprint planning, technical sequence diagrams) to deliver testable, production-ready features.
• Validated backend feature-engineering services in Python (Pandas), imputing missing values, normalizing engagement features to improve convergence of gradient-based optimization, engineering interaction-based features to ensure clean, reliable inputs for production systems.
• Evaluated linear (logistic regression) and nonlinear (decision tree, random forest) algorithms using cross- validation F1 and precision-recall trade-offs; integrated the optimal model into the product pipeline, improving recommendation accuracy by 18% and revenue by 10%.

Software Developer Intern - Soft scribble - Mumbai

(2023-05 - 2023-08)

• Scale data ingestion by engineering a custom API to process 100k+ data points monthly, significantly improving integration efficiency and backend scalability using SQL and JavaScript, contributing to the core development team as a full-stack contributor.
• Optimize system performance by refactoring queries, leading faster data retrieval, improved responsiveness for user-facing applications.
• Contributed to full Software Development Life Cycle (SDLC) from design to deployment, ensuring code quality through regular reviews testing.

MS - Information Science (Machine Learning) - University of Arizona (2024-08 - 2026-05)

Bachelor’s - Computer Science (AI & ML) - Mumbai University (2021-07 - 2024-05)