TuniPool

Shared GPU compute pool slashing AI cloud costs for Tunisian startups by 70%.

Score: 7.5/10TNMedium BuildReady to Spawn
Brand Colors

The Opportunity

Problem

Tunisian AI startups face high costs from international cloud services like AWS and Google Cloud, severely limiting their scalability and experimentation with AI tools.

Solution

TuniPool connects Tunisian AI startups to a shared pool of affordable local and regional GPUs, enabling cost-effective training and inference without international cloud bills. Users submit jobs via a simple dashboard, and the platform schedules them on underutilized hardware from local providers. This democratizes access to high-performance compute at fraction-of-AWS prices.

Target Audience

Tunisian AI startups building and scaling AI tools

Differentiator

Tunisian-first shared GPU marketplace with local billing in TND and partnerships with North African data centers for ultra-low latency.

Brand Voice

supportive

Features

Job Submission

must-have12h

Upload AI models/scripts and queue training/inference jobs

Dashboard Monitoring

must-have10h

Real-time job status, logs, and resource usage tracking

Auto-Scaling Queue

must-have15h

Intelligent job queuing on available GPUs with priority tiers

Cost Calculator

must-have8h

Pre-job cost estimator vs AWS/GCP benchmarks

TND Billing

must-have10h

Local currency invoicing and Stripe integration

API Access

must-have12h

REST API for programmatic job submission

Multi-Model Support

nice-to-have8h

PyTorch/TensorFlow/HuggingFace compatibility

Collaborative Workspaces

nice-to-have10h

Team sharing for projects

One-Click Deploy

nice-to-have12h

Instant inference endpoints

Advanced Analytics

future20h

Cost trends and optimization reports

Total Build Time: 117 hours

Database Schema

users

ColumnTypeNullable
iduuidNo
emailtextNo
stripe_idtextYes
credits_balanceintNo
created_attimestampNo

Relationships:

  • β€’ one-to-many with projects

projects

ColumnTypeNullable
iduuidNo
user_iduuidNo
nametextNo
gpu_typetextNo

Relationships:

  • β€’ foreign key to users.id, one-to-many with jobs

jobs

ColumnTypeNullable
iduuidNo
project_iduuidNo
statustextNo
gpu_hours_usedintYes
cost_tndintYes
completed_attimestampYes

Relationships:

  • β€’ foreign key to projects.id

API Endpoints

POST
/api/jobs

Submit new job

πŸ”’ Auth Required
GET
/api/jobs/:id

Get job status

πŸ”’ Auth Required
POST
/api/projects

Create project

πŸ”’ Auth Required
GET
/api/billing/credits

Check balance

πŸ”’ Auth Required
POST
/api/estimate

Cost estimation

πŸ”’ Auth Required

Tech Stack

Frontend
Next.js 14 + Tailwind + Shadcn/ui
Backend
Next.js API routes + Supabase Edge Functions
Database
Supabase Postgres
Auth
Supabase Auth
Payments
Stripe
Hosting
Vercel
Additional Tools
Resend (emails)Cron Jobs for job polling

Build Timeline

Week 1: Core auth and DB setup

20h
  • βœ“ User onboarding
  • βœ“ Basic DB schema
  • βœ“ Dashboard skeleton

Week 2: Project and job CRUD

25h
  • βœ“ Project creation
  • βœ“ Job submission UI/API
  • βœ“ Status polling

Week 3: GPU integration mock

25h
  • βœ“ Job queue simulation
  • βœ“ Monitoring dashboard
  • βœ“ Cost calc

Week 4: Billing and payments

20h
  • βœ“ Stripe integration
  • βœ“ TND pricing
  • βœ“ Credits system

Week 5: Polish and testing

15h
  • βœ“ User flows
  • βœ“ Error handling
  • βœ“ Mobile responsive

Week 6: Launch prep

10h
  • βœ“ Landing page
  • βœ“ SEO
  • βœ“ Analytics
Total Timeline: 6 weeks β€’ 125 hours

Pricing Tiers

Free

$0/mo

No priority, community GPUs only

  • βœ“10 GPU hours/month
  • βœ“Basic queue
  • βœ“1 project

Pro

$25/mo

Standard GPUs

  • βœ“200 GPU hours/month
  • βœ“Priority queue
  • βœ“Unlimited projects
  • βœ“API access

Enterprise

$99/mo

None

  • βœ“Unlimited hours (pay-per-use)
  • βœ“Dedicated GPUs
  • βœ“Custom SLAs
  • βœ“Team accounts

Revenue Projections

MonthUsersConversionMRRARR
Month 1504%$50$600
Month 63008%$1,920$23,040

Unit Economics

$40
CAC
$360
LTV
5%
Churn
85%
Margin
LTV:CAC Ratio: 9.0xExcellent!

Landing Page Copy

Cut AI Compute Costs by 70% – Tunisian GPUs for Tunisian Startups

Ditch AWS bills. Pool power with TuniPool for affordable, local AI training & inference.

Feature Highlights

βœ“70% cheaper than AWS
βœ“TND billing
βœ“Real-time monitoring
βœ“Easy API
βœ“Tunisian latency

Social Proof (Placeholders)

"'Saved us $2k/month!' – AI Startup Founder"
"'Finally affordable GPUs!' – Tunis Dev"

First Three Customers

Reach out to Tunisian AI communities on LinkedIn/Facebook groups like 'AI Tunisia' and 'Startups Tunisie'; offer free Pro tier for beta feedback; DM 20 founders from recent Techapoli events with personalized cost savings demo.

Launch Channels

Product Huntr/MachineLearningTwitter #AITunisiaLinkedIn Tunisia Tech

SEO Keywords

cheap GPU TunisiaAI compute TunisiaTunisian GPU rentalreduce AWS cost AIlocal AI hosting Tunisia

Competitive Analysis

RunPod

runpod.io
Pay-per-hour $0.2+
Strength

Global scale

Weakness

High costs for Tunisia, latency issues

Our Advantage

70% cheaper, local TND billing, Tunisia-optimized

Vast.ai

vast.ai
Marketplace variable
Strength

P2P cheap

Weakness

Unreliable, no local focus

Our Advantage

Curated Tunisian pool with SLAs

🏰 Moat Strategy

Network effects from growing shared pool; exclusive local data center partnerships.

⏰ Why Now?

AI boom in Tunisia post-2023 gov incentives; cloud costs spiking 50% YoY for startups.

Risks & Mitigation

technicalmedium severity

GPU supply shortages

Mitigation

Partner with multiple local providers; fallback to regional clouds

markethigh severity

Low adoption

Mitigation

Free tier + community outreach

executionmedium severity

Job orchestration bugs

Mitigation

Start with simulated jobs, rigorous testing

Validation Roadmap

pre-build7 days

Survey 50 Tunisian AI founders on Discord/LinkedIn

Success: 20+ express interest in beta

mvp14 days

Beta with 10 users

Success: 80% retention, positive NPS

launch7 days

PH launch

Success: 100 signups

Pivot Options

  • β†’General cloud hosting marketplace
  • β†’AI model registry only
  • β†’Enterprise consulting add-on

Quick Stats

Build Time
125h
Target MRR (6 mo)
$2,000
Market Size
$2.0M
Features
10
Database Tables
3
API Endpoints
5