Public Tender: Development of the AI-Powered WIN Gender Balance Tracker

Issuing Organisation: WAN-IFRA WIN

Tender Period: 10 December – 31 December (3 weeks)

Project Duration: 1 January 2026 – 30 June 2026

Submission Email: farah.wael@womeninnews.org

Background

WAN-IFRA WIN is a global programme working to advance gender equality in the media sector. WIN is commissioning the development of an AI-powered Gender Balance Tracker that will analyse gender representation, portrayal, roles, and bias in media content across text, images, audio, and video, in multiple languages.

The tracker will serve as a global evidence and accountability tool for newsrooms, journalists, editors, researchers, and advocacy organisations.

Objective

The objective of this tender is to design, develop, test, and deploy a fully functional AI-powered Gender Balance Tracker that:

• Analyses text, images, audio, and video
• Generates a Gender Balance Score based on quantitative and qualitative gender balance indicators including sources, and main characters
• Supports multilingual analysis
• Includes a user-friendly web interface with personal user accounts and data storage
• Is fully operational by 30 June 2026

Scope of Work

The selected service provider will be responsible for the full technical delivery of the platform, including:

A. AI & Data Processing

• Integration with OpenAI, Google Gemini, or equivalent AI models
• Text gender analysis using quantitative and qualitative indicators
• Image and photo gender analysis using computer vision
• Audio transcription and spoken-voice gender detection
• Video audio-track analysis with:
  • Speaking-time by gender
  • Role detection (host, guest, expert)
  • Screen-time and voice-time calculation
• Sentiment, framing, stereotype, and bias detection
• Gender-neutral language suggestions and loaded-phrase alerts

B. Gender Indicators & Outputs

The tracker must produce:

Quantitative Indicators

• Women as main characters
• Women as sources
• Women as authors or creators
• Mentions of women vs men
• Women as expert vs non-expert sources

Qualitative Indicators

• Role taxonomy and agency
• Sentiment and framing analysis
• Stereotype and bias detection
• Cultural and institutional tone analysis

Final Outputs

• Gender Balance Score (0–100)
• Automated summary with actionable recommendations
• Exportable reports

C. Multilingual Capability

The platform must support English, Arabic and Ukrainian. With language-specific datasets and aligned Gender Balance Scores.

D. Platform & User Interface

• User-friendly web interface
• Analysis by:
  • Single link
  • Batch/list of links
  • Direct text input
• Page and section scraping for non-paywalled websites
• User registration, login, and logout
• Personal user profiles
• Data storage and usage logs
• Admin dashboard with aggregated statistics
• Report export and download functionality
• Secure hosting and data protection

E. Testing & Handover

• Model accuracy testing
• Performance and security testing
• UX testing with real users
• Full system documentation
• User manual and technical documentation
• Final deployment and handover

Required Expertise

Bidders must demonstrate:

• Proven experience in:
  • Artificial Intelligence
  • Natural Language Processing (NLP)
  • Computer Vision
  • Speech-to-Text systems
• Experience integrating OpenAI, Gemini, or equivalent models
• Experience with multilingual NLP
• Secure backend and cloud infrastructure development
• Development of media monitoring or analytics platforms is a strong advantage

Proposal Submission Requirements

Interested bidders must submit:

1. Technical Proposal
2. Financial Proposal with a detailed budget
3. Company Profile including portfolio of similar projects

Submission Details

Proposals must be submitted by email to:

farah.wael@womeninnews.org

Deadline: 31 December 2025

Subject Line: WIN AI Gender Balance Tracker – Technical Tender

Late submissions will not be accepted.