Enterprise AI Architecture · Agentic Systems · Salesforce · Decision Intelligence
Building AI systems that turn enterprise context into evidence-backed decisions and governed action.
I’m an enterprise technology and transformation leader with 22 years in the Salesforce ecosystem, working at the intersection of enterprise architecture, business transformation and applied AI.
My central question is: How can we move AI beyond chat interfaces and isolated copilots into systems that understand business context, connect to enterprise tools, and help people make better decisions—with evidence, governance and human control?
This GitHub is my proof-of-work portfolio: original projects, practical experiments and reusable starting points exploring that question. I value understandable architecture, reproducible quality and useful outcomes over impressive demos alone.
|
Flagship · Enterprise growth and decision intelligence A Claude-powered platform for business analysis across accounts, sales, customers, competition, marketing, finance, meetings and decisions. It explores how specialist agents, business memory, deterministic calculations, evidence and approval gates can work together as a decision layer. Explore: Enterprise AI, agent orchestration, decision support, business memory, analytics and governed actions. |
Relationship analytics inside Salesforce Explores graph-based relationship intelligence and connected enterprise workflows, using algorithms to help surface patterns in business relationships. Explore: Salesforce architecture, graph analytics and relationship intelligence. |
|
Published in the Anthropic directory · v1.1.0 A multi-stage workflow for research, drafting, skeptical editing, copy editing and fact-checking—designed to make writing more deliberate, source-aware and reviewable than one-shot generation. Available through: Claude Code, Cowork and Claude apps. Explore: Claude skills, research-led writing and content quality. |
⚙️ VendorPulseProcurement workflows in Salesforce A Salesforce project focused on purchase-order renewals, renewal status, contract values, reporting and spreadsheet-based data movement. Explore: Salesforce, Lightning Web Components, Apex and operational workflows. |
| Principle | How I apply it |
|---|---|
| Evidence over guesses | Make sources, assumptions and inputs inspectable. |
| Outcomes over theatre | Solve a concrete workflow or decision problem. |
| Governed autonomy | Keep consequential actions reviewable and approval-driven. |
| Deterministic where it matters | Use code for calculations and explicit contracts for critical outputs. |
| Quality is part of the product | Treat tests, evaluations, documentation and review as core work. |
| Reusable by design | Make systems understandable, modular and adaptable. |
I publish perspectives on Salesforce architecture, enterprise delivery and technology transformation through Doctor Salesforce on Medium.
- Salesforce Headless: A Strategic Repositioning, Not a Revolution — examining the architecture narrative around headless Salesforce.
- Salesforce Project Survival Manual — 100 Unspoken Truths — lessons on delivery, technical realities, scope and governance.
- The Ethical Implications of Using AI and Machine Learning in Healthcare — perspectives on privacy, bias and transparency.
Read the ideas. Explore the code. Adapt what works. I aim to connect practical technology commentary with projects people can inspect and build on.
- Enterprise AI and decision intelligence — connecting business context, evidence, analysis and action.
- Agentic workflows — designing repeatable systems around specialized tasks, not just prompts.
- MCP and enterprise integration — connecting AI to tools and business systems responsibly.
- Salesforce and enterprise architecture — applying architecture and integration thinking to operational needs.
- Knowledge systems — making context, research and decisions reusable.
- 💼 LinkedIn: linkedin.com/in/sumitmattey
- 🎥 YouTube: AI Guy
- ✍️ Doctor Salesforce: Medium publication
- 💻 GitHub: github.com/mattey2026
- 🔬 ORCID: 0009-0000-6304-894X
If a repository helps with a problem you are solving, read its README, check its current status and license, then fork and adapt what works. Constructive feedback and contributions are welcome through issues and pull requests.
Build thoughtfully. Share openly. Make AI useful.
Enterprise AI · Agentic systems · Salesforce architecture · MCP · Decision intelligence