Tako Sandbox ·

If an entrepreneur "adopted" this product today, here's the realistic math.

Fermi summary

If you land 14 paying teams at $1k/mo by month 12, that's $168k ARR - call it a 9% shot, meaning expected value is deeply negative year 1 and this only makes sense if you believe you can ride the AI agent wave past year 2.

Market size (TAM)

$180.0M

~12,000 engineering teams globally actively executing untrusted code in AI agents × $1,200/mo average infrastructure spend = ~$172M ARR addressable today, growing fast but still early

Year-1 ARR range

$42k - $480k

midpoint $168k

Gross margin

62%

Investment to production

$85k

Dev: $35k for multi-tenant isolation hardening, SDK polish, and billing/metering. Security audit: $18k for third-party container escape and

Probability of success

P(reaching mid case in 12 months)

Expected take-home Y1

$-75000

probability-weighted, after investment

Go-to-market motion

Developer-led: post on HN/AI communities → free tier hook → upsell AI startups and enterprise AI teams via usage-based pricing when their sandbox usage exceeds free limits.

Key risks

E2B already owns this category with open-source traction, a funded team, and an established community - differentiation must be extremely sharp (compliance posture, latency, language support) to justify switching cost
The 'zero blast radius' security guarantee is a liability, not just a feature - one publicized container escape or privilege escalation will terminate the product, so the security audit spend is non-optional and ongoing
Usage-based pricing creates a margin trap: the customers who stress-test sandboxes most (large agentic workflows, coding assistants) are also the ones who burn your compute budget fastest, compressing margins exactly when revenue grows

Generated by the Wishdeal Factory financial-analysis agent. Numbers are honest Fermi estimates, not guarantees. Real outcomes depend on the operator. The studio is bullish on the engineering quality, agnostic on the business outcome.