CJ • Portfolio

Chaitanya Shrikrishna Joshi

Senior Automation & AI Solution Engineer • Power Platform Architect • Data & AI Engineering

Automation and AI specialist with 3+ years of U.S. experience designing and delivering enterprise-grade, secure, and scalable solutions across Power Platform, Azure AI/ML, and RPA. Proven record at Emerson's Global Financial Services CoE (Fortune 500) leading multi-million-dollar process automation across finance, data, and compliance. Translate business challenges into high-value architectures, governance frameworks, and AI-enabled applications. Skilled in U.S. stakeholder engagement, cross-functional delivery, and measurable cost/time savings.

Power PlatformAzure OpenAIRAG PythonSQL/T-SQLUiPath AlteryxOCR

Highlighted Projects

More by CJ

🏠 Mortgage Document Classification

Featured

AI-powered document processing pipeline simulating Alteryx + OCR workflows. Classifies mortgage documents (appraisals, income verification, bank statements, tax returns) with 95%+ accuracy.

  • OCR integration simulation with realistic document templates
  • Multi-class classification with confidence scoring
  • Real-time analytics dashboard with business KPIs
  • Enterprise-ready architecture for 10K+ documents/day

Pricing & Margin Audit

FastAPI

Audit quotes vs. pricebooks, flag RED/YELLOW/GREEN deals, and auto-generate exec summaries via HF Inference.

Commission Forecast

Time Series

Forecast incentives from pipeline and attainment; configurable assumptions and payout curves.

Lead Agent

Agents

Scores leads, prioritizes outreach, and drafts contextual emails with tight CRM integrations.

PolicyGuard AI

RAG

Ask GDPR/CCPA/SOX questions over your docs with RAG and auditable citations.

Netflix Medallion Analytics

dbt • DuckDB • Power BI

Bronze→Silver→Gold lakehouse with dbt on DuckDB. Gold marts power a Power BI report. The live Space lets you run the dbt pipeline in-browser and explore the Gold layer.

  • Bronze: raw seed; Silver: typed/cleaned/exploded; Gold: genre, release_year, country marts
  • One-click dbt build in the Space; query DuckDB directly from the UI

🏠 Mortgage Document Classification - Technical Deep Dive

🔧 Technical Architecture

  • OCR Integration: Simulates Alteryx + Python (pytesseract, OpenCV) workflows for text extraction
  • ML Classification: Hybrid rule-based + ML approach with confidence scoring (90-95% accuracy)
  • Document Types: Appraisal Reports, Income Verification, Bank Statements, Tax Returns, Credit Reports
  • Real-time Analytics: Processing metrics, performance trends, business KPIs
  • Scalable Design: Cloud-ready architecture for enterprise deployment (10K+ docs/day)

💼 Business Impact

  • Time Savings: 96.7% reduction (15 min → 30 sec per document)
  • Cost Reduction: $732K annual savings potential
  • Accuracy: 95%+ classification accuracy with confidence scoring
  • Scalability: Handle peak processing loads automatically

🏗️ Implementation Approach

Phase 1: OCR Simulation & Classification Logic

Built realistic document templates and classification rules

Phase 2: Streamlit Dashboard & Analytics

Real-time processing metrics and business intelligence

Phase 3: Enterprise Architecture Design

Scalable cloud deployment with Alteryx integration

🎯 Design Decisions

  • Simulation Strategy: Focus on ML logic over OCR API integration
  • Portfolio Accessibility: Demo works without external dependencies
  • Production Ready: Architecture designed for easy OCR integration
  • Industry Focus: Mortgage-specific document understanding

Selected Enterprise Projects

Automated Mortgage Document Classification (Portfolio)

Enterprise-grade document processing pipeline with OCR + ML classification; demonstrates Alteryx integration patterns and 95%+ accuracy for financial document workflows.

PolicyGuard AI (Power Apps + Azure OpenAI)

RAG + Azure OpenAI in Power Apps for compliance reviews; reduced U.S. policy analysis time by 50%.

Natural Language SQL Agent

LLM-based SQL generator with Power BI connector; reduced analytics backlog by 60%.

Power Platform Governance Hub

CoE governance model with lifecycle controls, role matrix, and security compliance audits across U.S. orgs.

AI Invoice Reconciliation Tool

UiPath + Power Automate + OCR to reconcile invoice data; eliminated ~4 FTEs equivalent effort.

Netflix Data Lakehouse (dbt + DuckDB)

Medallion architecture implementation with Bronze→Silver→Gold layers; integrated Power BI for executive dashboards.

Experience

Intelligent Automation RPA, Data & Power Platform Engineer — Emerson Electric Co. (St. Louis, MO)

Jan 2022 – Jun 2024

Architected and delivered 20+ enterprise-scale automation & AI solutions for U.S. finance operations, enabling standardization, audit compliance, and measurable savings.

  • Dataverse migration for a $100M+ billing platform (D365 F&O, Oracle ERP) → −$100K/yr support costs; improved SOX audit readiness.
  • Python + OCR (Tesseract, OpenCV) for invoice processing → +50% classification accuracy; −400+ hours/month manual work.
  • LangChain-based RAG with Azure OpenAI for document Q&A → −30% analyst research time.
  • Power Platform governance (CoE Toolkit, secure connectors, DLP) for citizen developer enablement.
  • ALM CI/CD with Azure DevOps for Power Platform → −25% deployment defects.
  • Mentored U.S. interns; led rollouts across NA HQ and U.S. business units.

Business/Data Analyst Intern — University at Buffalo (Buffalo, NY)

Sep 2024 – Jun 2025
  • DirectLake Power BI dashboards for academic performance; integrated ADF pipelines and Dataverse schema.
  • Automated reporting with Power Automate + Python → +20% cadence/accuracy.
  • SharePoint-integrated portals for faster faculty data access and decisions.

Product Development Lead (Part-Time) — Machinery Monitoring Systems LLC (Remote, US)

Jan 2025 – Jun 2025
  • Real-time IoT telemetry pipeline (PySpark, AWS S3, Lambda) → predictive maintenance models; −30% downtime.
  • Azure ML APIs embedded in diagnostics dashboards → −40% issue resolution time.
  • Stakeholder workshops with U.S. engineering to scale PoCs into production.

Skills

Architecture & Strategy

Enterprise automation design, AI/ML integration, ALM, CI/CD, CoE enablement, governance frameworks, document processing pipelines.

Platforms

Power Apps, Power Automate, Power BI, Copilot Studio, Dataverse; Azure Logic Apps, Functions, OpenAI, AI Search, Document Intelligence; Alteryx Designer.

Programming & Integration

Python (Pandas, NumPy, Scikit-Learn, PyTorch), SQL/T-SQL, REST APIs, Custom Connectors, C#, JavaScript, PL/SQL, dbt, DuckDB.

Automation & AI

UiPath (REFramework, Orchestrator), Power Automate Desktop, RAG pipelines, AI Builder, LangChain, LLM integration, OCR (Tesseract, OpenCV), document classification.

Data & Analytics

Azure Data Factory, Databricks, Snowflake, Streaming Analytics, Power BI semantic modeling, DAX, KPI dashboards, Medallion architecture, data lakehouse patterns.

Governance & Security

Role-based access, DLP policies, API security, lifecycle governance, audit compliance, financial document processing compliance.

Soft Skills: Cross-functional collaboration, stakeholder engagement, mentoring, Agile/Scrum, technical architecture design, enterprise solution delivery

Research & Publications

Certifications & Training

  • UiPath Certified Professional — Process Automation Developer
  • Python Complete Developer (Udemy)
  • MS Fabric Data Engineer (In progress)
  • Alteryx Designer Core Certification (In progress)
  • OpenAI Integration, LangChain, LlamaIndex, LLMOps
  • Multi-AI Agent Systems with crewAI
  • Applied AI for Managers — University at Buffalo
  • OCR & Document Processing (Tesseract, OpenCV)

Education

MS, Management Information Systems (STEM) — University at Buffalo, SUNY
Jun 2025
DBMS, Predictive Analytics, Cloud Data Warehousing, Systems Analysis & Design, Data Visualization, Applied AI for Managers, IT Project Management, Experiential IT Projects, Gen AI Consulting, Tech & Innovation Mgmt.
MCA — Vishwakarma Institute of Technology, University of Pune
Apr 2022
B.Sc. Computer Science — University of Pune
Apr 2019

Contact

Open to roles in Data Engineering, Automation & AI (GenAI), Power Platform architecture, and Document Processing/OCR solutions. Remote and hybrid across the U.S.

Tip: the "Highlighted Projects" section links to live demos hosted on Hugging Face Spaces. Try the 🏠 Mortgage Classification demo for an interactive experience!