Social video, turned intosearchable political-influence intelligence.
WITI — Who Influences The Influencer — a proof-of-concept dashboard converting TikTok and social video into transcribed, searchable, AI-analysed political influence data.
The situation
Platforms like TikTok shape public discourse quickly, but the influence networks, messaging patterns and content strategies underneath are slow and expensive to analyse manually at any scale.
MMA wanted researchers to move from ad hoc observation to a structured view of political narratives, influence patterns and the strategies shaping public opinion.
What we built
An AWS-based processing workflow: content from selected TikTok accounts is uploaded, converted and pre-processed, transcribed into text, then AI-analysed for themes, narratives, sentiment and influencer-level patterns.
A dashboard presenting the results in a searchable, filterable format, so researchers can review content by account, topic, narrative or time period.
What changed
Video content that used to demand manual video-by-video review became structured civic intelligence that can be searched, compared and analysed — with repeatable analysis in place of one-off observation.
MMA gained a reusable technical foundation for future political influence and information-integrity monitoring.
As a proof of concept, it de-risks a larger future platform by showing funders and stakeholders that social video can be transformed into analysable data.
“From video-by-video review to structured civic intelligence.”