Omowaye Joshua Oluwaseun

Cybersecurity | 🤖 AI | ☁️ Cloud Security
🛠️ Building AI-powered automation systems
📊 Learning in public | 🚀 Future Security Engineer

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 23: stats per import batch.

Batch 1: 3 NEW, 0% converted.

One lead → CONVERTED → 33.33%. Live SQL, not hardcoded.

Bad ID → None. No crash.

Gap: new timestamp can still clone Jane. Stats won’t save you.

#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 22: import batches are visible in the CLI.

Batch 1 → 3 leads. Bad ID 999 → no crash.

Gap: new timestamp still clones Jane (ID 11, now also 14). History ≠ lock.

#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 21: same cleanup batch can’t import twice.

First: 3 in. Replay: 0 in, 3 skipped. Still 3 leads.

Lock is the batch key, not “email unique.”

Gap: new timestamp = new batch. Same CSV tomorrow can clone Jane.

#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 20: CLEAN rows finally hit SQLite.

3 imported. Status NEW. CREATED activity. CLI sees them.

Broke: python -m app.crm_cli → missing error_handler / database. Needed app. imports.

Gap: run twice = duplicate leads. Skip-by-email is Day 21.

#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 19: each cleanup gets its own folder.

output/Client/cleanup_timestamp/
CSV opens first. Missing file = error, no empty folder.

6 files per run. Clients don’t mix.

CLEAN still isn’t in SQLite.

#Python #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

@DataSynthGen Thanks! Honestly, the biggest surprise was how fast small issues tank the score. A few missing fields and duplicate groups can drag down CRM health quickly.

It made me realize data cleaning isn’t just about fixing errors it’s about making the database trustworthy for the team

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 18: CRM health score.

Validity 40 / uniqueness 35 / completeness 25.

Messy file → 81% GOOD. Clean file → 100%. Empty/missing CSV errors.

Dupes = groups, not pairs÷2.

#Python #DataQuality #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 17: fuzzy duplicates. RapidFuzz, not “AI.”
Same name ≠ same customer.

John Adewale vs Jon Adewale: name 95%, company 86%, email 76%, phone 43% → POSSIBLE DUPLICATE, not merge.

HIGH = same email/phone. MEDIUM = human review. No auto-delete
#Python #RapidFuzz #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 16: bulk CRM cleanup.

7-row CSV → 3 CLEAN / 2 REVIEW / 2 DUPLICATE.

Same email+phone = dupe. Flagged, not deleted.

0803… → [phone]. Original kept.

Dirty in. Three files out. DB stays clean.

#Python #CSV #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 15: status changes now write history.

Before: overwrite the column. Old value gone. Now: save old → new, then update.

CLI kept CONVERTED → LOST. Dashboard un-counted it.

Snapshot ≠ audit trail

Next: first outreach auto NEW → CONTACTED

#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 14: added status to a live SQLite CRM without wiping data.

PRAGMA, then ALTER only if missing. Started lead_activities

CONTACTED → CONVERTED. 1/7 = 14.29%.
Status = snapshot. History = audit trail.

Next: first outreach auto NEW → CONTACTED.
#Python #SQLite #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

Day 13 — My CRM dashboard said I converted 100% of leads.

But status breakdown said: CONVERTED: 0.

The bug? One SQL query was counting every lead.

Fixed it with the correct WHERE status = 'CONVERTED'.

Lesson: metrics are only as honest as the SQL behind them.

#Python #SQLite

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 12 — AI CRM Automation

Improved lead management.

✅ Lead status tracking
✅ Status updates & search
✅ Sales pipeline stages
✅ Fixed database mapping
Workflow: Lead → AI → Database → CRM → Status
#Python #AI #CRM #Automation #BuildInPublic #100DaysOfCode #NVIDIA

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 11 — AI CRM Automation

Added AI-powered lead analysis.

✅ NVIDIA AI
✅ Intent & product detection
✅ Lead scoring
✅ Automated summaries

Workflow: Lead → AI → Database → CRM

#Python #AI #CRM #Automation #BuildInPublic #100DaysOfCode

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 10 — AI CRM Automation

Improved CRM reliability with error handling.

✅ Protected AI & database operations
✅ Added user-friendly errors
✅ Tested error handling
Workflow: Raw Lead → Validation → AI → Database → CRM
#Python #AI #CRM #Automation #BuildInPublic

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 9 — AI CRM Automation

Integrated validation into the CRM.

✅ Block invalid leads
✅ Validate before AI & database
✅ Tested full pipeline

Workflow: Raw Lead → Validation → AI → Database → CRM
#Python #AI #CRM #Automation #BuildInPublic #100DaysOfCode

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 8 — AI CRM Automation

Built a validation layer for cleaner CRM data.

✅ Lead validation
✅ Email & phone checks
✅ Priority & score va
✅ Automated tests
Workflow: Raw Lead → Validation → AI → Database → CRM
#Python #AI #CRM #Automation #BuildInPublic ##100DaysOfCode

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 7 — AI CRM Automation

Connected AI processing directly to the CRM.

✅ Lead creation
✅ Data cleaning
✅ Duplicate detection
✅ AI analysis & scoring
✅ SQLite storage
Workflow: Raw Lead → AI → Database → CRM
#Python #AI #CRM #Automation #BuildInPublic #100DaysOfCode

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 6 — AI CRM Automation

Built CRM lead management.

✅ View lead by ID
✅ Update priority & score
✅ Delete leads safely
✅ Delete confirmation
✅ Full CRUD with SQLite
Create → Read → Update → Delete
#Python #AI #CRM #BuildInPublic #100DaysOfCode #NVIDIA

[Omowaye Joshua Oluwaseun @CtrlSEUN]

🚀 Day 5 — AI CRM Automation

Built the interactive CRM CLI.

✅ Menu-driven interface
✅ View all leads
✅ Search by email
✅ Filter by priority
✅ Lead
✅ Connected to SQLite
Raw Lead → AI Analysis → Database → CRM CLI
#Python #AI #CRM #BuildInPublic #100DaysOfCode #NVIDIA