Diag ‐ delete - magicplatforms/ai-workflows GitHub Wiki
Below is a ready‑to‑paste Azure DevOps Wiki page. It includes a table of contents, Azure DevOps note‑callouts, and nine Mermaid sequence diagrams comparing a “typical day” before and after AI across major business functions.
How to use
- Make sure the Preview feature “Render Mermaid diagrams in Wiki and Markdown” is enabled for your organization / project.
- Copy everything between the horizontal rules into a new Wiki page.
AI‑Driven Workflow Transformation
[[TOC]]
[!NOTE] Mermaid diagrams render automatically once the page is saved or previewed (no additional extensions required).
Sales – Before vs After AI
sequenceDiagram
title Sales – Before vs After AI
participant SDR
participant AE
participant LinkedIn
participant Prospect
participant CRM
participant AI as AI_Agent
participant Calendar
participant Coach as Win_Coach
alt **Before AI**
SDR->>LinkedIn: Scrape leads
SDR->>Prospect: Draft & send emails
SDR->>CRM: Log calls
AE->>CRM: Update opps nightly
else **After AI**
AI->>IntentDB: Mine intent signals
AI->>Prospect: Auto‑personalise outreach
AI->>Calendar: Schedule call
AE->>Coach: Realtime win‑probability tips
CRM-->>AE: Forecast auto‑roll‑up
end
Human Resources – Before vs After AI
sequenceDiagram
title HR – Before vs After AI
participant Recruiter
participant Candidate
participant HRBP
participant ResumeDB
participant Calendar
participant AI as GenAI_Ranker
participant Chatbot
participant Analytics as HR_Analytics
alt **Before AI**
Recruiter->>ResumeDB: Screen 200 résumés
Recruiter->>Candidate: Phone screens
Recruiter->>Calendar: Block interview slots
HRBP->>HR_Analytics: Build turnover deck monthly
else **After AI**
AI->>ResumeDB: Rank résumés
Chatbot->>Candidate: Run 1st interview
Calendar->>Panel: Auto‑book interviews
Analytics-->>HRBP: Live attrition heat‑map
HRBP->>Analytics: Push‑button stay‑bonus scenario
end
Legal – Before vs After AI
sequenceDiagram
title Legal – Before vs After AI
participant Associate
participant Paralegal
participant Westlaw
participant Client
participant LLM as Research_LLM
participant ContractAI
alt **Before AI**
Associate->>Westlaw: Hunt precedent
Associate->>Client: Draft memo & redline contract
Paralegal->>Docs: Bates‑stamp discovery
else **After AI**
LLM->>Westlaw: Search & summarise precedent
ContractAI->>Client: Mark risk clauses & draft NDA
Paralegal->>Docs: Validate AI‑coded discovery
Paralegal->>Team: Prep strategy brief
end
Finance / Banking – Before vs After AI
sequenceDiagram
title Finance / Banking – Before vs After AI
participant Analyst
participant Excel
participant Compliance
participant AI as MarketAI
participant Surveillance as Surveillance_AI
alt **Before AI**
Analyst->>DataFeeds: Copy data to Excel
Analyst->>Excel: Build valuation model
Analyst->>Note: Write 8‑page report
Compliance->>Trades: Hand‑check overnight
else **After AI**
AI->>FilingsDB: Ingest filings & generate model
AI-->>Note: Draft equity research note
Analyst->>AI: Stress‑test assumptions
Surveillance-->>Trades: Pre‑clear in seconds
end
Education – Before vs After AI
sequenceDiagram
title Education – Before vs After AI
participant Teacher
participant Student
participant Parent
participant Grader
participant PlanAI as Lesson_Generator
participant TutorAI
participant Analytics
alt **Before AI**
Teacher->>Plan: Craft lesson plan
Teacher->>Grader: Manually grade papers
Teacher->>Parent: Send emails
else **After AI**
PlanAI-->>Teacher: Differentiated lesson plan
TutorAI->>Student: Instant feedback
Analytics-->>Teacher: Performance dashboards
Teacher->>Class: Socratic discussion
Teacher-->>Parent: 30‑sec video update
end
Technology / Software Dev – Before vs After AI
sequenceDiagram
title Software Dev – Before vs After AI
participant Engineer
participant Repo
participant CI
participant Copilot
participant LLM as BuildHelp_LLM
alt **Before AI**
Engineer->>Repo: Design API & write boilerplate
Repo->>CI: Wait for build & tests
CI-->>Engineer: Code review feedback
else **After AI**
Copilot-->>Repo: 60–80 % scaffolding & autotests
Engineer->>Code: Focus on edge cases & architecture
CI->>LLM: Build errors
LLM-->>Engineer: Explain fixes in chat
CI-->>Prod: Release cadence halves
end
Healthcare – Before vs After AI
sequenceDiagram
title Healthcare – Before vs After AI
participant Physician
participant Patient
participant Guidelines
participant Coder
participant ScribeAI as Ambient_Scribe
participant ClinicalLLM
alt **Before AI**
Physician->>EHR: Type notes
Physician->>Patient: Dictate letters
Physician->>Guidelines: Search care guidelines
Coder->>Claim: Assign CPT codes
else **After AI**
ScribeAI-->>EHR: Auto‑generate note
ClinicalLLM->>Physician: Diagnostic suggestions & orders
Coder->>Claim: Audit AI‑assigned codes
Physician->>Patient: Spend extra time counselling
end
Supply‑Chain / Ops – Before vs After AI
sequenceDiagram
title Supply‑Chain / Ops – Before vs After AI
participant Planner
participant ERP
participant Supplier
participant ML as Demand_Model
participant GenAI
participant Twin as Digital_Twin
participant Router as Routing_Engine
alt **Before AI**
Planner->>ERP: Export data & tweak forecast
Planner->>Supplier: Call for updates
Dispatcher->>Routes: Build route sheet
else **After AI**
ML-->>Planner: Demand forecast
GenAI->>Supplier: Draft PO & negotiation email
Twin-->>Planner: Network simulation
Router-->>Carrier: Issue loads autonomously
Planner->>Exceptions: Handle only anomalies
end
Insurance (P&C / Life) – Before vs After AI
sequenceDiagram
title Insurance – Before vs After AI
participant Adjuster
participant Underwriter
participant Claim
participant CV as CV_Model
participant GenAI
participant FraudAI
alt **Before AI**
Adjuster->>Claim: Read claim & request photos
Adjuster->>PartsDB: Price parts
Adjuster->>Report: Write assessment
Underwriter->>Tables: Crunch risk tables
else **After AI**
CV->>ClaimPhotos: Score damage from images
GenAI-->>Claim: Draft settlement
Underwriter->>RiskAI: Review AI risk score
Underwriter->>RiskAI: Tune appetite
FraudAI-->>Claim: Flag anomalies in real‑time
end
End of page