👋 Start a project Blog posts, videos, and unfiltered experience-sharing from running 20+ ventures in parallel. What worked, what broke, what I'd do differently. No agency theatre — just receipts from the trenches.
The economics of parallel AI startups. Which 3 are working (ORBIX, NOBBYO, Voxly). What I'd cut. Portfolio thinking vs project thinking.
The 4-8 week AI MVP playbook used to ship Voxly, ORBIX, and NOBBYO. Exact stack, week-by-week milestones, the traps to avoid. By a founder who's done it 20 times.
Real numbers: $250k year-1 CTO cost vs an equity-aligned founder-operator partnership. When each makes sense. Includes a decision matrix.
250M+ Bangla speakers. English-trained AI fails them at every turn. Why Bangla NLP is the niche AI opportunity for 2026 — and the 5 hard problems waiting to be solved.
Vercel + Supabase + Anthropic free credits + Cloudflare = $0/month AI MVP. Exact stack, what to upgrade and when. Real cost breakdown.
How a 6-person Bangladesh-based team out-ships UK agencies for 40% of the cost. NDA, IP, DPA playbook for international clients.
How niches were found for ORBIX, NOBBYO, Voxly, Pannakhata and 14 other ventures. 3 techniques that work, 5 dead-ends to avoid.
AI / Web / Brand / Marketing / Video / BizDev / Social Media — which services every founder needs in year 1, what's seasonal, what to skip.
The 6 roles that beat a 12-person agency. Hiring sequence, handoff playbook, tool stack, and compensation. By a founder who runs this team daily.
Year-1 cost, speed, equity, and risk side-by-side. When co-build wins. When agency wins. When CTO wins. Real founder examples.
One weekly note. Build logs, hard-won lessons, and the occasional free template. No spam, no funnels. Unsubscribe in one click.
I usually reply to Whatsapp in 20 mins.