Built and launched Soma AI MVP in 3 weeks. Automated Meta ads workflow, reduced setup time by ~90%, and enabled faster, scalable campaign launches for teams.
Performance teams launching bulk Meta ads spend hours repeating the same setup tasks: uploading creatives, mapping placements, naming, duplicating across campaigns, etc. This slows experimentation and increases errors. Soma set out to compress Meta ad setup from hours to minutes.
2. Goal of the MVP
Validate that growth teams would adopt Soma for daily Meta ad ops and that it would measurably reduce ad launch time.
3. My Role & Stack
Owned product engineering end to end.
HLD + technical feasibility
Researched and integrated Meta Marketing APIs
Built frontend and backend
Deployed and managed infra
Stack: NextJs, Node, Meta Marketing API, MongoDB, Vercel
4. Approach & Key Decisions
Focused only on critical value moments rather than “nice to haves”
Import creatives directly from Google Drive
Persistent naming and settings
Auto placement mapping and grouping
Multi campaign launch
Designed workflows around how growth teams already operate
Optimized for speed, reliability, and zero-blocker adoption
Simplified UX so teams could move their workflow in a single day
5. Outcome & Impact
Adopted by teams like SouledStore, GrowthSchool, Blinkist, Kutumb
Enabled teams to save ~6 hours per week per 100 ads launched (from ~4 minutes per ad down to ~24 seconds)
Supported every Meta ad format including Advantage+ and Lead Gen
Strong validation that Meta ad ops is a real, painful, monetizable problem