SHAMS Network Planner

An AI assistant that turns connectivity needs into practical network plans, device recommendations, cost estimates, and ready-to-send proposals for homes, businesses, and underserved communities.

Comment

Inspiration

Inspiration

In many underserved communities, getting internet access is only the first challenge. People also struggle to choose the right routers, antennas, power equipment, coverage design, and network capacity.

As the founder of Shams Phone in Sudan, I regularly see homes, small businesses, and local network operators purchase incompatible equipment or build networks without a clear technical plan. Professional network consultation can be expensive or unavailable, especially outside major cities.

SHAMS Network Planner was inspired by the need to make practical network planning easier, faster, and more accessible.

What it does

SHAMS Network Planner is an AI-powered assistant that turns basic connectivity requirements into a practical network plan.

The user provides information such as:

  • Number of users
  • Coverage area
  • Internet source, such as Starlink
  • Available budget
  • Existing equipment
  • Indoor or outdoor deployment

The system then generates:

  • A recommended network topology
  • Suggested routers, access points, antennas, and supporting equipment
  • Bandwidth and capacity recommendations
  • Power and PoE compatibility warnings
  • Coverage assumptions and installation guidance
  • A preliminary cost estimate
  • A customer-ready proposal
  • A concise WhatsApp message that can be shared with the customer

The goal is not to replace a certified network engineer. The system clearly displays assumptions and recommends professional verification for complex or high-risk deployments.

How we are building it

During OpenAI Build Week, the project is being developed as a mobile-first web application using Codex and GPT-5.6.

Codex is being used to help:

  • Design the application architecture
  • Build the user interface
  • Implement validation rules
  • Write and review code
  • Create tests
  • Identify and repair errors
  • Prepare technical documentation

GPT-5.6 converts the user’s requirements into structured recommendations and clear explanations.

The project combines AI reasoning with deterministic safety rules. Important issues such as voltage, PoE compatibility, unrealistic coverage expectations, and excessive user capacity are checked before recommendations are displayed.

Challenges

The main challenge is converting real-world field experience into clear and testable software rules.

Other challenges include:

  • Preventing unsupported or unsafe recommendations
  • Explaining uncertainty instead of presenting guesses as facts
  • Designing for users with limited technical knowledge
  • Creating a useful experience on mobile phones
  • Supporting communities with limited connectivity and limited budgets
  • Building a functional prototype within the Build Week deadline

What we learned

This project shows that AI works best when it is combined with practical domain knowledge and strict validation.

A good network recommendation must not only sound intelligent. It must account for capacity, distance, power, compatibility, budget, and real deployment conditions.

We also learned that clearly explaining assumptions is essential for building trust.

What's next

Future versions could include:

  • Interactive network diagrams
  • Map-based coverage planning
  • Local equipment and pricing catalogs
  • PDF proposal export
  • Arabic and English interfaces
  • Offline or low-bandwidth support
  • Inventory integration
  • Saved customer projects
  • Follow-up installation checklists

Our long-term goal is to make reliable connectivity planning accessible to homes, businesses, and underserved communities wherever professional technical support is difficult to reach.

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for SHAMS Network Planner

Built With

Share this project:

Updates