Swanand Barsawade smiling at Pacific Park

Swanand Barsawade

Data & AI PractitionerProcess IntelligenceEnterprise Automation

I am a data and AI practitioner focused on process intelligence and intelligent automation. I design and ship enterprise solutions using Celonis, UiPath and machine learning to improve process performance and unlock measurable business value. I also build production grade ML and GenAI applications, including LLM, RAG and agentic workflows, with close attention to reliability and maintainability. I lead programs end to end, from data pipeline design and architecture to rollout, adoption and value tracking. I am known for clear execution, thoughtful technical judgment and outcomes that teams can trust. I bring a product mindset that aligns technical execution with customer value and business priorities.

Experience

I have led enterprise programs across process mining, UiPath aligned automation, machine learning, experimentation and BI delivery. I build scalable data and AI systems on strong engineering foundations, with reliable pipelines and architecture patterns that hold up in production. I translate complex business questions into production ready systems with measurable impact.

Projects

These projects reflect my approach as a practitioner: start with business outcomes, build the right analytical or AI solution and engineer for scale. I communicate clear decision ready results.

Enterprise & AI Projects

Local-First RAG Pipeline

Production document QA system built on LangGraph, Ollama, and embedded Qdrant. Reduced end-to-end latency from 28.7s to under 10s via model quantization and elimination of Docker dependencies. Hybrid BM25 and vector retrieval, cross-encoder reranking, Plan-Execute-Verify agent architecture.

OpenClaw Foundry

14-layer zero-trust AI agent enclave on bare metal. Four specialized agents with inotifywait-based prefix routing, persistent memory, Vault KV v2 secret management, nftables egress isolation, NeMo guardrails, AIDE integrity monitoring, and S3 WORM audit trail.

Academic & Research Projects

Finding Ideal Location for a New Restaurant

Analyzed Yelp data to identify high potential expansion zones for restaurants. Built clustering, sentiment analysis and classification workflows, then mapped recommendations using Google Maps APIs.

Micro Credit Loan Defaulter

Built a credit risk classification pipeline using Logistic Regression, SVM and tree based models to identify likely defaulters from repayment history and financial indicators.

Sales Data Analysis of Walmart Stores

Developed a PySpark based forecasting workflow for Walmart department sales, combining feature analysis and regression modeling for more reliable planning.

EDA of Avocado Dataset

Performed end to end EDA and price prediction on avocado market data using SVR, Decision Tree and Random Forest regressors.

Predicting Customer Retention

Built churn prediction models on telecom customer data to identify at risk users and support targeted retention strategies.

Implementing Genetic Algorithm for Feature Optimization

Implemented a genetic algorithm for knapsack style optimization and evaluated convergence and solution quality across configurations.

Skills, Architecture & Leadership

I combine deep technical capability with delivery leadership. My toolkit spans data science, process mining, intelligent automation, AI engineering, data engineering and analytics architecture. I operate with high standards for technical rigor, reliability and measurable outcomes.

Python Celonis UiPath LangChain LangGraph Apache Spark SQL Tableau Azure Docker RAG LLMs Process Mining

Get in touch

Connect to discuss data, AI, process intelligence and product building ideas.

Résumé