Instant AI fact-checking with West African context
West African journalists are overwhelmed by AI-powered disinformation that erodes public trust and threatens social cohesion.
FactFalcon is a mobile app that lets journalists upload text, images, or videos for immediate AI analysis trained on regional data. It delivers authenticity scores, detects AI generation patterns, and surfaces relevant local historical and cultural context via RAG to counter disinformation effectively. This enables faster, more trustworthy reporting in high-stakes environments.
West African journalists and media professionals covering sensitive political and social topics
Hyper-localized RAG database built from West African news archives, election records, and cultural datasets that global tools completely miss.
supportive
Upload or paste content for instant AI authenticity and disinformation risk scoring
Retrieves culturally relevant local context using vector search
Analyzes images and videos for synthetic generation markers
One-tap professional PDF reports with sources and confidence levels
Cached models and data for field use without internet
Rates quoted sources based on historical reliability in region
Securely share analyses with newsroom colleagues
Notifications about rising disinformation topics in your beat
Personal dashboard showing disinformation exposure by topic
Support for Wolof, Hausa, and French dialects
| Column | Type | Nullable |
|---|---|---|
| id | uuid | No |
| text | No | |
| full_name | text | Yes |
| media_outlet | text | Yes |
| regions | text | Yes |
| created_at | timestamp | No |
Relationships:
| Column | Type | Nullable |
|---|---|---|
| id | uuid | No |
| user_id | uuid | No |
| input_hash | text | No |
| score | int | No |
| is_disinfo | bool | No |
| ai_model_used | text | Yes |
| created_at | timestamp | No |
Relationships:
| Column | Type | Nullable |
|---|---|---|
| id | uuid | No |
| analysis_id | uuid | No |
| context_text | text | No |
| source_url | text | Yes |
| relevance | int | No |
| embedding | text | Yes |
Relationships:
| Column | Type | Nullable |
|---|---|---|
| id | uuid | No |
| user_id | uuid | No |
| tier | text | No |
| paystack_ref | text | Yes |
| status | text | No |
| current_period_ends | timestamp | Yes |
Relationships:
/api/analyzeSubmit content for fact-checking and context retrieval
/api/historyFetch user's previous analyses with pagination
/api/context/searchVector similarity search for local context
/api/reportsGenerate and return PDF verification report
/api/webhook/paystackProcess Paystack subscription events
20 checks per month
None
Up to 10 seats
| Month | Users | Conversion | MRR | ARR |
|---|---|---|---|---|
| Month 1 | 180 | 9% | $405 | $4,860 |
| Month 6 | 1,450 | 17% | $6,162 | $73,944 |
AI that actually understands West Africa. Real-time verification with local context for journalists who refuse to be fooled.
Partner with the Ghana Journalists Association and Nigerian Union of Journalists to offer free Professional accounts to 25 active members each in exchange for case studies. Run targeted LinkedIn outreach to senior reporters at Premium Times, JoyNews, and RFI. Present at the next West Africa Media Excellence Awards to generate early buzz and signups.
Extensive unreliable site database
Western-centric with minimal West African coverage
Purpose-built RAG for regional politics, culture and languages
Strong AI text detection
No contextual understanding or local knowledge
Combines detection with culturally relevant explanations
Proprietary West African disinformation corpus that grows with anonymized journalist usage, creating a flywheel of increasingly accurate regional intelligence.
Generative AI tools became widely accessible in 2023 exactly as multiple West African nations head into high-stakes elections, creating urgent demand for specialized defense tools.
AI model inaccuracies in low-resource languages
Combine multiple models with clear confidence indicators and human review prompts
Journalists wary of new tech tools
Heavy focus on co-creation with early users and transparent methodology
Potential liability for incorrect assessments
Prominent disclaimers, 'for assistance only' language, and insurance
Success: Confirmed willingness to pay $20+ monthly
Success: 65% weekly active users and positive NPS
Success: 500 downloads and $1,200 MRR in 30 days
Other validated startup ideas you might find interesting
Never miss TechCabal articles again—search and recover 404 pages instantly.
Your personal vault for TechCabal links—auto-recovers 404s forever.
AI revives lost TechCabal pages—summarize, rewrite, recover.
Get warm enterprise intros in days, not months for AI founders.
Auto-generate interactive enterprise demos that close deals faster.
AI crafts winning enterprise proposals that land meetings instantly.