Semantic Search The Influship API uses AI-powered semantic search meaning behind your query.
How It Works
Your query is converted into a semantic embeddingCreator profiles are matched based on content, style, and audienceResults are ranked by relevance with AI-generated explanations
What We Match On Signal Description Content Topics What creators post about Content Style How they create (educational, entertaining, aesthetic) Audience Signals Engagement patterns and follower demographics Platform Behavior Posting frequency, format preferences
Writing Effective Queries
Good Queries
“fitness creators who focus on home workouts”
“tech reviewers with highly engaged audiences”
“fashion influencers promoting sustainable brands”
Less Effective Queries
“influencer” (too vague)
“good creator” (no topic context)
“someone popular” (no semantic meaning)
Query Tips Tip Example Be specific about niche ”vegan cooking” not just “food” Mention audience traits ”with engaged Gen Z audience” Include content style ”educational tutorials” vs “entertainment” Add platform context ”Instagram Reels creators”
Understanding Results Each search result includes:
{ "id" : "550e8400-..." , "name" : "Sarah Johnson" , "match" : { "score" : 0.92 , "reasons" : [ "Strong focus on sustainable fashion content" , "Consistently high engagement on outfit posts" , "Active collaboration with eco-friendly brands" ] } }
interface SearchResult { id : string ; name : string ; match : { score : number ; // 0-1 reasons : string []; }; } const results = await client . creators . search ({ query: 'sustainable fashion creators' }); results . items . forEach ( creator => { console . log ( ` ${ creator . name } : ${ creator . match . score } ` ); creator . match . reasons . forEach ( r => console . log ( ` - ${ r } ` )); });
Score Interpretation Combining Search with Filters Semantic search works best when combined with hard filters:
cURL
TypeScript SDK
Python
curl -X POST https://api.influship.com/v1/search \ -H 'X-API-Key: YOUR_API_KEY' \ -H 'Content-Type: application/json' \ -d '{ "query": "sustainable fashion creators", "filters": { "platform_filters": [{ "platform": "instagram", "min_followers": 10000, "max_followers": 500000, "min_engagement_rate": 2.0 }] }, "limit": 25 }'
const results = await client . creators . search ({ query: 'sustainable fashion creators' , filters: { platform_filters: [{ platform: 'instagram' , min_followers: 10000 , max_followers: 500000 , min_engagement_rate: 2.0 }] }, limit: 25 });
results = requests.post( 'https://api.influship.com/v1/search' , headers = { 'X-API-Key' : API_KEY }, json = { 'query' : 'sustainable fashion creators' , 'filters' : { 'platform_filters' : [{ 'platform' : 'instagram' , 'min_followers' : 10000 , 'max_followers' : 500000 , 'min_engagement_rate' : 2.0 }] }, 'limit' : 25 } ).json()
This finds creators who:
Semantically match “sustainable fashion”
Have 10K-500K Instagram followers
Have at least 2% engagement rate
Relevant Profile Selection For multi-platform creators, the API returns a relevant_profile field showing which profile best matched your query:
{ "id" : "550e8400-..." , "relevant_profile" : { "platform" : "instagram" , "username" : "eco_sarah" , "match_reason" : "Primary platform for fashion content" } }
This helps when a creator is on multiple platforms but your query is platform-specific.
Show Advanced: How Semantic Embeddings Work
The API uses state-of-the-art transformer models to convert text into high-dimensional vectors. These vectors capture semantic meaning, so “sustainable fashion” and “eco-friendly clothing” are mathematically close in the embedding space, even though they share no keywords. This enables the AI to understand intent, synonyms, and context that keyword search would miss.
