How srutip04/influencerdiscovery works
InfluencerDiscovery is a collection of design and sample artifacts for a large-scale influencer discovery system, not an implemented application. Its documented scope includes architecture and design materials, example Airflow DAGs, API and data schemas, an embeddings-ranking demo, and infrastructure and cost artifacts. The DAGs define schedules but contain no tasks in the excerpts. Since no runtime calls or data transfers are shown, the graph groups the artifacts by role without asserting connections between them.
Components 17
Architecture and design
- Architecture diagram
architecture/architecture_diagram.mmd - Diagram notes
architecture/diagram_legend_and_notes.md - Design write-up
design-doc/design_writeup.md
Workflow examples
- Ingestion and backfill
airflow/dags/ingestion_and_backfill_dag.py - Enrichment pipeline
airflow/dags/enrichment_pipeline_dag.py - Identity resolution
airflow/dags/identity_resolution_dag.py - Monitoring alerts
airflow/dags/monitoring_alerts_dag.py
Data and API
- API specification
api/openapi.yaml - Creator profile schema
schemas/creator_profile.sql - Platform profiles
schemas/platform_profiles.sql - Identity graph
schemas/identity_graph_schema.sql - Feature store
schemas/feature_store.sql - Creator features
schemas/dynamodb_creator_features.json - Vector index metadata
schemas/vector_index_metadata.sql
Demo and infrastructure
- Embedding ranking demo
notebook/embeddings_ranking_demo.ipynb - Infrastructure snippets
infra/terraform/README_TF.md - Cost model
infra/cost_model/cost_model.md