Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I’m Nishit Mistry, a data scientist and data engineering professional currently working on building practical, production-minded ML and AI solutions that translate messy real-world data into decision support. My recent work includes developing a soybean yield prediction and variety-ranking pipeline using Python, SQL, and cross-validation strategies, and building RAG-based tools that replace static dashboards with faster, breeder-friendly recommendations. I also enjoy leading applied NLP projects—from extracting reimbursement rules from complex insurance contracts to deploying RAG-powered chatbots for billing support—while collaborating with teams to deliver measurable outcomes like faster turnaround times and improved precision. I’m especially interested in making model evaluation align with how stakeholders actually choose results, and I bring a strong foundation in statistics and machine learning alongside hands-on experience in shipping reliable data pipelines.…I’m Nishit Mistry, a data scientist and data engineering professional currently working on building practical, production-minded ML and AI solutions that translate messy real-world data into decision support. My recent work includes developing a soybean yield prediction and variety-ranking pipeline using Python, SQL, and cross-validation strategies, and building RAG-based tools that replace static dashboards with faster, breeder-friendly recommendations. I also enjoy leading applied NLP projects—from extracting reimbursement rules from complex insurance contracts to deploying RAG-powered chatbots for billing support—while collaborating with teams to deliver measurable outcomes like faster turnaround times and improved precision. I’m especially interested in making model evaluation align with how stakeholders actually choose results, and I bring a strong foundation in statistics and machine learning alongside hands-on experience in shipping reliable data pipelines.

Nishit Mistry

Data Scientist, Data Engineer, Programmer, +2





I’m Nishit Mistry, a data scientist and data engineering professional currently working on building practical, production-minded ML and AI solutions that translate messy real-world data into decision support. My recent work includes developing a soybean yield prediction and variety-ranking pipeline using Python, SQL, and cross-validation strategies, and building RAG-based tools that replace static dashboards with faster, breeder-friendly recommendations. I also enjoy leading applied NLP projects—from extracting reimbursement rules from complex insurance contracts to deploying RAG-powered chatbots for billing support—while collaborating with teams to deliver measurable outcomes like faster turnaround times and improved precision. I’m especially interested in making model evaluation align with how stakeholders actually choose results, and I bring a strong foundation in statistics and machine learning alongside hands-on experience in shipping reliable data pipelines.…I’m Nishit Mistry, a data scientist and data engineering professional currently working on building practical, production-minded ML and AI solutions that translate messy real-world data into decision support. My recent work includes developing a soybean yield prediction and variety-ranking pipeline using Python, SQL, and cross-validation strategies, and building RAG-based tools that replace static dashboards with faster, breeder-friendly recommendations. I also enjoy leading applied NLP projects—from extracting reimbursement rules from complex insurance contracts to deploying RAG-powered chatbots for billing support—while collaborating with teams to deliver measurable outcomes like faster turnaround times and improved precision. I’m especially interested in making model evaluation align with how stakeholders actually choose results, and I bring a strong foundation in statistics and machine learning alongside hands-on experience in shipping reliable data pipelines.

Available to hire

I’m Nishit Mistry, a data scientist and data engineering professional currently working on building practical, production-minded ML and AI solutions that translate messy real-world data into decision support. My recent work includes developing a soybean yield prediction and variety-ranking pipeline using Python, SQL, and cross-validation strategies, and building RAG-based tools that replace static dashboards with faster, breeder-friendly recommendations.

I also enjoy leading applied NLP projects—from extracting reimbursement rules from complex insurance contracts to deploying RAG-powered chatbots for billing support—while collaborating with teams to deliver measurable outcomes like faster turnaround times and improved precision. I’m especially interested in making model evaluation align with how stakeholders actually choose results, and I bring a strong foundation in statistics and machine learning alongside hands-on experience in shipping reliable data pipelines.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

Work Experience

Data Scientist at Soybean Innovation Lab – University of Illinois Urbana-Champaign

January 1, 2026 - Present

Engineered a soybean yield prediction and variety-ranking pipeline using Python (pandas, scikit-learn), SQL, Power BI, and GroupKFold CV across 500K+ multi-country trial records, reframing evaluation from RMSE to Spearman rank correlation and adding Top-10 accuracy/precision aligned to breeder selection. Architected a RAG-based decision-support tool over multi-country soybean trial data using Python and GPT-4o, replacing static Power BI dashboards and reducing decision time by 45 minutes per variety.

Lead Consultant, Data Science at Business Intelligence Group – University of Illinois Urbana-Champaign

August 1, 2025 - May 1, 2026

Built a multi-layer NLP system (regex, entity recognition, semantic parsing) to extract reimbursement rules from complex insurance contracts, achieving 90% extraction consistency across an 80-contract sample. Deployed a billing-support AI chatbot using Streamlit, the OpenAI API, and RAG for a client’s patient billing portal, translating client scoping into a working prototype within 6 weeks. Led a team of 5 consultants across two semester-long engagements, owning technical architecture for NLP/RAG and driving milestone delivery and client cadence.

Data Science Intern & Co-op at Sherwin-Williams

May 1, 2025 - October 1, 2025

Streamlined an end-to-end NLP analytics pipeline over 22,343 customer complaints, enabling R&D scientists to answer recurring product/quality questions 6x faster (25 minutes down to 4 minutes). Deployed NLP workflows including text preprocessing, classification, and LLM-based summarization to transform unstructured complaint text into 14 structured categories consumed by R&D analysis. Operationalized the pipeline in Dataiku with idempotent replays and backfilled 12 months of historical data to eliminate duplicate loads, and delivered Snowflake-backed Tableau dashboards that reduced weekly ad-hoc analysis requests.

Assistant Manager, Data Engineering at Larsen & Toubro Financial Services

July 1, 2022 - June 1, 2024

Launched 12+ Salesforce Flow journeys (e.g., CIBIL, payments, communications, stop banking) with automated case creation, reducing manual triage by 40% and improving first-touch accuracy by 20%. Coordinated Salesforce ETL/ELT work (validations, transforms, case creation) on daily/weekly schedules to achieve 98% on-time success and cut incident tickets by 30% after launch. Developed internal REST API services for loan data and secure document retrieval (SOA/NOC/repayment schedules), reducing response time by 25% and eliminating ticket-based data pulls. Consolidated 1M+ rows (7–10 years) into Salesforce dashboards keyed on loan and case IDs, replacing fragile Excel exports for cross-product trend analysis in wealth management.