โ† Back to Portfolio
๐Ÿ›ฑ Case Study 01 ยท Lead Quality & Outreach

Mystery Shopping AI Automation System

A quality-assurance outreach engine: every prospect address is verified at the DNS + SMTP protocol level before a single email is spent, then campaigns fire from four inboxes on a human-paced schedule.

๐Ÿ“ธ Built running live

$ build_optimized_mystery_verifier.py
[DNS] resolving MX records ... cached per domain
[SMTP] RCPT TO probe โ†’ catch-all detection
[RESULT] 5,000 raw โ†’ 2,847 verified โ†’ 1,9xx deliverable-whales
Verification waterfall โ€” this run happens for every lead
$ SCHEDULE_MYSTERYSHOPPING_CAMPAIGN.bat
Inboxes : owais@ ยท hirabaig@ ยท wes@unseenresearch ยท wes@wesms
Leads   : 22 / 22 / 22 / 21
Stagger : ~11 min between sends per inbox
Window  : Mon 18:30โ€“22:30 IST
4-inbox staggered send that looks human, because it is paced like a human

โš™๏ธ How the pipeline runs

Lead Pool
CSV โ†’ master DB
โ†’
DNS Layer
MX records ยท provider detect ยท catch-all flag
โ†’
SMTP Layer
toxic-domain kill ยท RCPT probe ยท domain cache
โ†’
Verifier
concurrent workers, cached domains
โ†’
Zoho Scheduler
4 inboxes ยท 11-min stagger ยท time-windowed
โ†’
Campaign Status Monitor
per-account progress

๐Ÿง— The engineering moat

๐Ÿ“Š Client-facing outcome

4inboxes orchestrated
~11 minhuman-paced stagger
100%of sends verified pre-flight
0paid verification APIs

Drop a raw list in. Get back a campaign that only emails real, deliverable, human-operated inboxes โ€” deliverable-checked recipients โ€” from multiple sender identities, on a schedule that mimics manual outreach.