Click here
Vanchhit Khare
16X Salesforce Certified Architect | 9+ Years Experience | Leading with Emotion, Building with Precision
System Architect. Researcher. Educator. Musician.
About Me
Building Tech That Actually Helps People
Hi, I'm Vanchhit Khare, a System Architect with over nine years of experience. I believe technology should empower people, not overshadow them, and this philosophy fuels my work with platforms like Salesforce, delivering real difference to organizations.
My passion for impactful technology extends to research; I've published at several IEEE conferences. My 'Water Saviour Systems' project earned a Best Paper award for solving issues in society and will be published by Springer, reflecting my commitment to solutions that positively impact communities.
Beyond my professional roles, I teach students about AI to nonprofits and also teach LinkedIn tricks and tips on TECHBRIDGE, and unwind by making music with Ableton Live or Logic Pro X. The true reward for me is knowing something I've built or created has made someone's day easier or more harmonious.
🎯 Core Expertise & Certifications
Top Skills
  • Agentforce Implementation
  • Database Management (SQL)
  • Experience Cloud
  • Financial Services Cloud
  • Generative AI Integration
  • GitLab & CI/CD Processes
  • Knowledge Site Management
  • OmniStudio Development
  • Power Apps & Microsoft Azure
  • Process Automation & Flows
  • Sales Cloud & Service Cloud
  • Salesforce Platform Architecture
Key Certifications
  • Development Lifecycle and Deployment Architect
  • Salesforce Certified Application Architect
  • Salesforce Certified Platform Data Architect
  • Salesforce Certified Platform Identity and Access Management Architect
  • Salesforce Certified Platform Integration Architect
  • Salesforce Certified Platform Sharing and Visibility Architect
  • Salesforce Certified System Architect
🏦 Work Experience
Assistant Vice President – M&T Bank | Dec 2024 – Present
Secure SSO Integration
PowerApps Solutions
Azure Integration
Knowledge Site Implementation (8,000+ users + Genesis chatbot)
Automated Provisioning (JIT scripts + AD Groups)
Release Management (multiple sandboxes)
Content Architecture (Knowledge articles)
Experience Cloud Development
Database & Analytics (SQL schemas + dashboards)
Vice President – M&T Bank | Dec 2022 – Dec 2024
CRM Implementation (FSC)
Security & Compliance
System Integration (GitLab/CI/CD)
OmniStudio Innovation (FlexCards + OmniScripts)
Business Systems Administrator – Vizient | Jul 2020 – Nov 2022
Release Management Leadership (VSCODE/XML)
Business Analysis & Configuration (100+ profiles)
Scrum Collaboration & Administration
🌱 Early Career
1
MFour Mobile Research – Salesforce Administrator (Dec 2019 – Jul 2020)
  • UI/UX
  • Workflow
  • Process Builder
  • Flows
  • Hubspot
  • Ring Central
  • AirCall
  • Clarizen
  • Cloudingo
  • Inverto360
2
Quantum Electronic Payments – Salesforce Administrator (Jul 2019 – Dec 2019)
  • Sales & service cloud
  • Reports
  • Dashboards
  • Liaison between users/vendors/dev teams
3
Ashunya Inc – Cloud Software Engineer Intern (Aug 2018 – Dec 2018)
  • Development requirements
  • Admin functions
  • Change management
📚 Published Research & Innovation
My commitment to advancing technology extends beyond daily work into meaningful research that addresses real-world challenges.
A DESIGN FOR FUSE IC INBUILT MEMORY SAVING ELEMENT – memory optimization in integrated circuits
MATNOB – MULTI ACCESS TO NETWORK ORIENTED BROWSERS – browser architecture for multiple network access
AUTOMATIC SWITCH CUM FUSE IC FOR STAND ALONE SYSTEMS – smart protection for standalone systems
Optical Navigation System for Robotics Application – optical sensors for robotic movement
ADVANCED LASER TECHNOLOGY FOR CURING CANCER AFFECTED MELANOMA CELL – laser cancer treatment research
📬 Connect With Me
Ready to transform your organization's technology landscape? I'm here to help with Salesforce solutions, system optimization, and innovative approaches.
Workshop Proposal // Spring 2026
Building a Habit Tracker iOS App with AI in 60 Minutes
Five live projects. No slides. I build everything in front of the class using Claude Code and Claude.ai, starting from a blank folder and ending with a working SwiftUI app on screen.
1
1 Hour
2
Mixed Level Students
3
iOS App Finale
4
5 Live Projects
What I'll actually do for 60 minutes
I open a blank folder on my laptop. The projector shows my terminal. Over the next hour, I build five things in front of the class: a formatted README, an architecture diagram, a project config that teaches AI my coding rules, a live connection to external tools, and finally a full SwiftUI habit-tracker app. All using AI.
Each project adds one concept. By Project 5, students already understand how AI reads context and follows instructions. So when it starts writing Swift code, they can follow along instead of just staring.
Fair warning: I've rehearsed this, but AI doesn't always cooperate. Last time I ran a similar demo, Claude generated a SwiftUI layout that compiled but looked like a ransom note. I fixed it live, and honestly that became the most useful part. Students saw how to catch and correct AI mistakes, which is the real skill.
The 5 Projects
Project 1: AI Document Generator (8 min)
I type one sentence into Claude.ai: "Write a README for a habit-tracker iOS app." Then we watch it produce a formatted document with headings, tables, and code blocks. Students see that AI isn't a vending machine — you talk to it, push back, iterate. Tools: Claude.ai, Markdown.
  • Markdown file structure
  • Prompt engineering basics
  • AI as a writing collaborator
  • Iterative refinement with AI
Project 2: AI Flowchart & Diagram Maker (8 min)
I describe the habit-tracker app's screens in plain English. Claude generates a Mermaid flowchart: user opens app, checks login state, routes to dashboard or onboarding. This is how engineers actually plan before writing code. We'll use this exact diagram as the blueprint for Project 5. Tools: Claude.ai, Mermaid Diagrams
  • Flowchart design (Mermaid syntax)
  • System architecture thinking
  • User journey mapping
  • AI for visual planning
Project 3: Custom AI Skills & CLAUDE.md (10 min)
I open the terminal and create a CLAUDE.md file that tells Claude our rules: use SwiftUI, 4-space indentation, no force unwraps. Then I build a reusable Skill — a saved prompt that runs like a slash command. The point: AI can be taught YOUR rules. Without this file, Claude just guesses. Tools: Claude Code, CLAUDE.md, Custom Skills
  • CLAUDE.md project configuration
  • Custom Skills (slash commands)
  • AI personalization & rules
  • Terminal-based AI (Claude Code CLI)
  • Prompt templates & reusability
Project 4: MCP Server Integration (10 min)
I edit one JSON file (.mcp.json) and suddenly Claude Code can search the web, read local files, and query a database. MCP (Model Context Protocol) is how you give AI access to the outside world — like USB ports for AI. I'll connect a live search server so students can see Claude pull fresh data mid-conversation. Tools: Claude Code, MCP Servers, .mcp.json
  • MCP (Model Context Protocol)
  • Server configuration
  • AI tool integration
  • Extending AI capabilities
  • Real-world AI actions (not just chat)
Project 5: Build an iOS App with AI (20 min)
Everything comes together. I tell Claude Code: "Build the habit-tracker app from our Project 2 flowchart." It reads the CLAUDE.md rules from Project 3, uses the MCP connection from Project 4, and starts writing Swift files. Students watch in real-time as it creates views, models, and navigation. The flowchart from Project 2 becomes the actual app. Tools: Claude Code, SwiftUI, Skills, MCP, CLAUDE.md
  • SwiftUI basics (built by AI)
  • iOS app architecture
  • AI as a full-stack developer
  • Agentic coding
  • Code review of AI output
  • When to trust vs. verify AI code
60-Minute Agenda
1
P1: Documents — 0:00–0:08
2
P2: Flowcharts — 0:08–0:16
3
P3: Skills & CLAUDE.md — 0:16–0:26
4
P4: MCP Servers — 0:26–0:36
5
P5: iOS App — 0:36–0:56
Note: Minutes 0:00–0:02 = Quick intro. Minutes 0:56–1:00 = Q&A and resources.
Why This Works
Progressive complexity
Project 1 is a simple doc. By Project 5, students are building apps. Each step adds ONE new concept so nobody gets lost.
Zero slides, all live
Every concept is demonstrated in real-time. Students see AI working, not bullet points ABOUT AI working.
Callback architecture
Project 5 uses the flowchart from Project 2, the CLAUDE.md from Project 3, and the MCP from Project 4. Everything connects.
Mixed-level friendly
Beginners learn "AI can write docs." Intermediate students learn "AI can be configured." Everyone gets value.
Real-world relevant
MCP servers, Skills, and CLAUDE.md are what professional developers use daily in 2026. Students learn industry tools.
Safety and ethics built in
I talk about the DataTalks.Club incident where AI deleted 2.5 years of production data. Students learn that "AI built it" doesn't mean "AI built it right."
What We Need
From the University
  • Projector / large screen for live coding
  • Stable Wi-Fi connection
  • Room with 30–100 student capacity
  • HDMI / USB-C display connection
I Will Bring
  • MacBook with Xcode + Claude Code installed
  • Pre-configured MCP servers (backup)
  • All 5 projects pre-tested (fallback demos)
  • Resource links handout for students
Students Need
  • Nothing! This is a live demo workshop
  • Optional: laptop to follow along
  • Optional: free Claude.ai account
  • Curiosity and questions encouraged
About the Speaker
Vanchhit Khare is a Solutions Developer at M&T Bank with 10+ years of software development experience. He finished his Master's in Computer Science from CSUF in 2019, where he worked with Professor Falkow. Coming back to speak to CSUF students feels personal.
Outside of his day job, he is a published researcher and active peer reviewer. He has authored 8 research papers, completed 40+ peer reviews across IEEE, Springer, Elsevier, MDPI, and IGI Global, and spoken at 5+ conferences. His most recent publication, 'What AI Agent Failures Look Like from the Inside, and Why Article 73 Misses Them' (Tech Policy Press, April 2026), analyzes how multi-agent coordination failures fall through the EU AI Act's incident reporting framework.
He builds multi-agent systems for enterprise environments and works daily with Claude Code, custom Skills, MCP servers, and the same tools he'll demo in this workshop.
10+
Years in Software
8
Research Papers
40+
Peer Reviews
5+
Conferences
Vanchhit Khare · Workshop Proposal for CSUF // Spring 2026 · 5 Projects · 1 Hour · Zero Slides · 100% Live