Hello!
I'm Sakshi Patel, an aspiring GenAI Engineer and Data Specialist, building intelligent, data-driven systems that automate complex processes with innovative technologies.
Get in touch sakshipatel2598@gmail.com
Recent graduate with a Master's in Information Technology and Managementfrom The University of Texas at Dallas. I have a strong background in generative AI and data engineering, with a passion for building scalable, data-driven systems
Currently, I work as an AI Engineer at WorldLink, where I design and develop stateful, agentic AI systems using LangGraph. My work involves building Neo4j-based supply-chain graphs, implementing route optimization and risk modeling using graph algorithms, and developing AI-driven pipelines for tariff computation and decision automation.
Previously, I worked as a GenAI Engineer Intern at Insight Global, where I built an AI-powered Statement of Work generator using a multi-agent architecture and developed a Retrieval-Augmented Generation (RAG) system to improve document accuracy through vector-based retrieval. Before that, I gained strong industry experience at LTIMindtree, engineering high-volume data pipelines and building analytics dashboards using Python, SQL, ETL tools, and Tableau.
Drafting a Statement of Work (SOW) is a vital part of business and legal projects. It outlines key details like deliverables, timelines, responsibilities, and legal terms. However, creating these documents is often a slow and complex process. This paper introduces a new AI-driven automation system that makes the entire SOW drafting process faster, easier, and more accurate. Instead of relying completely on humans, the system uses three intelligent components or 'agents' that each handle a part of the job.
This solution shows how artificial intelligence can be used to support legal and business professionals by taking care of routine work and helping them focus on more important decisions. It's a step toward making legal processes smarter, faster, and more reliable.
Read more...
- Python
- JavaScript
- Typescript
- SQL
- Java
- R
- React.js
- Node.js
- Flask
- GraphQL
- Tableau
- Power BI
- MongoDB
- PostgreSQL
- Snowflake
- AWS
- Azure
- Seeburger BIS
- LangGraph & LangChain
- RAG
- Multi-Agent Systems
- Prompt and Context Engineering
- Vector Search
- Architected production-grade, stateful LangGraph agentic systems with token budgeting and controlled context windowing, ensuring stable LLM responses and per-session chat persistence in SQL.
- Improved LLM response quality and latency using structured prompt engineering and context compaction, reducing token usage while preserving task-critical context.
- Built and optimized Neo4j-based supply-chain graphs using Cypher and graph algorithms (Dijkstra, k-shortest path) to power route optimization and risk modeling, improving route-planning efficiency by 25%.
- Modeled supply-chain domain logic across product classification, country-of-origin, and tariff rules in Neo4j, enabling fast graph traversal and reducing rule-resolution time by 40%.
- Built AI-driven tariff computation pipelines automating duty calculations across HTS classification, origin, weight, and cost, reducing manual analysis and improving accuracy by 30%.
- Developed an AI-powered SOW generator using LangGraph with multi-agent orchestration, automating drafting, validation, and feedback loops to reduce manual effort by 30%.
- Built a Retrieval-Augmented Generation (RAG) system leveraging LLMs and NLP-driven automation, improving document accuracy by 25% and incorporating vector embeddings stored in PostgreSQL (pgvector).
- Created automated ETL pipelines using Python and SQL to extract, clean, and transform large-scale driver training data, enhancing data integrity and reducing processing time by 35%.
- • Deployed the project on Azure, utilizing Azure Compute for scalable processing and Azure PostgreSQL for secure, high-performance data storage alongside integration with Azure OpenAI for LLM inference.
- Mentored 50+ students on AWS, Hadoop/Spark, Tableau, Power BI, and Python for real-world data analytics projects.
- Developed curriculum and provided feedback to enhance analytical skills
- Engineered and deployed EDI-based payment integrations on Seeburger BIS, enabling migration of large-scale payment systems by processing 100M+ transactional records achieving 99.99% system uptime and 94% transformation accuracy.
- Designed and implemented SQL-driven data ingestion and validation pipelines across SIT and UAT environments, automating database writes, file-level validations, and transaction checks, reducing manual intervention by 50%.
- Developed Java-based validation overlays and assertion logic, enforcing strict file-structure and instruction-level compliance. automated CI/CD deployments via Jenkins, and validated end-to-end flows using TOSCA.
- Authored High-Level Design (HLD) and Low-Level Design (LLD) documents, defining system architecture, data flows, and integration touchpoints. collaborated closely with business analysts to translate business requirements.
- Created Power BI and Tableau dashboards highlighting user retention and growth, increasing consultation-based revenue by 25%.
- Developed SQL queries and AWS Athena pipelines for real-time metrics and program performance analysis.
- Automated Excel-based stakeholder reports, reducing repetitive tasks by 50%
- Built and optimized SQL pipelines to extract and transform REST API data for business reporting, decreasing query response time by 40%.
- Designed and deployed Tableau dashboards for automated insights, reducing manual reporting by 20%.
Engineered an end-to-end clinical analytics pipeline and machine learning models (Logistic Regression, Random Forest, XGBoost) to predict clinical outcomes on MIMIC ICU data, processing millions of structured and time-series patient records (labs, vitals, diagnoses, medications)
Built Hadoop/Spark infrastructure to process 10M+ records, reducing latency by 30%. And created React-based dashboard integrated with Tableau to track driver behavior and incident risk zones, reducing fleet accidents by 25%.
Analyzed 3M+ grocery transactions with Python & SQL, improving cross-sell strategy by 15% & forecast accuracy by 25%. And Visualized key insights using Tableau dashboards to support strategic planning and boost retention by 18%.
Integrated an internal knowledge management chatbot using LangChain and OpenAI, enabling employees to retrieve insights from 1,000+ documents with 85% retrieval accuracy.