Hi, I am Shruti Misra
AI Engineer Building Production-Grade Agentic Systems
I help organizations move AI from experimentation to production.
My work focuses on multi-agent systems, enterprise RAG, evaluation
frameworks, and AI platforms that transform enterprise data into
intelligent products, automated workflows, and measurable business
outcomes.
Ph.D. in Electrical & Computer Engineering (University of Washington)
| Databricks Certified | Snowflake Certified
Projects
Enterprise AI: 85% Faster HR Case Resolution with Multi-Agent Intelligence
As part of the team at Aimpoint, I build an agent that not only achieved 85%+ faster case resolution, but also 96% monthly active user adoption and 94%+ user satisfaction/response quality.
- Python
- Snowflake
Procurement Intelligence in Under 1 Minute
Built an AI-powered procurement intelligence solution for Simfoni that transformed complex spend and supplier data into actionable recommendations in under 1 minute. Developed as a production-ready Snowflake Native App using agentic AI to accelerate sourcing decisions, identify cost-saving opportunities, and improve procurement efficiency.
- Python
- Snowflake
- Streamlit
Enterprise RAG for 70M+ Documents with 90%+ Classification Accuracy
This project is centered on building an evaluation-driven RAG system that classified over 70 million documents, achieved 90%+ classification accuracy, delivered context-aware answers in under 30 seconds, and included an MLflow-based evaluation framework.
- LangGraph
- MlFlow
- Python
- Databricks
Sparse Transformer on FPGA
Microsoft internship project where I developed and assessed the feasibility of sparse computations for BERT run of FPGAs
- Python
- ModelSim
Interpretable AI-Decision Trees
In this project, I assessed the interpretability of decision trees by using them to predict whether a patient has diabetes based on their profile.
- Python
- Sklearn
Predicting Breast Cancer in a Timely Manner
This project was part of the Women in Data Science (WiDS) Datathon 2024, where I competed in a team of two other people. The goal was to predict whether a patient would receive their cancer diagnosis within 90 days. Our best model's diagnosis time accuracy was 79.3% as compared to 80.1% of the winning model
- Python
- Sklearn
Blogs & Videos
Building an AI-Ready Semantic Layer in Snowflake Cortex
In this demo, I show how Snowflake Cortex Analyst can answer natural-language business questions more reliably when data has been enriched with clear business meaning.
- Databricks
- Tool-Calling Agent
- Business Intelligence
Inventory Management Made Easy Using AI/BI Genie
An AI-powered inventory management solution using BI Genie to automate stock forecasting and streamline supply chain decisions.
- Databricks
- Text2SQL
- Business Intelligence
Smarter Maps, Smarter Markets: Geospatial GenAI for Real Estate Investment
From Maps to Meaning: Building Agentic AI Systems with Geospatial Context
- Databricks
- Tool-Calling Agent
- Business Intelligence
Inventory Management Made Easy Using AI/BI Genie
An AI-powered inventory management solution using BI Genie to automate stock forecasting and streamline supply chain decisions.
- Databricks
- Text2SQL
- Business Intelligence
Accelerating Financial Data Analysis Using Snowflake's Cortex Analyst
Leveraging Snowflake Cortex Analyst to accelerate financial data analysis through natural language queries and AI-driven insights.
- Snowflake
- Text2SQL
- Financial Analysis
Leveraging Snowflake's Cortex Search to Build RAG Systems
Exploring how Snowflake Cortex Search enables scalable Retrieval-Augmented Generation (RAG) systems for enterprise-grade semantic search and LLM workflows.
- Snowflake
- Retrieval Augmented Generation (RAG)
AI Engineering
- LangGraph
- LangChain
- RAG
- Agentic AI
- Context Engineering
- Agent Evaluation
Platforms
- Databricks
- Snowflake
- AWS
Backend Engineering
- Python
- FastAPI
- Docker
- Git
Observability & Evaluation
- MLflow
- TruLens
- OpenTelemetry
- Online Evaluation
- Offline Evaluation