Influencer Marketing Platform

I design and develop the ML/AI layer as the core of the product—the primary driver of business value, not just an add-on to the application.

ML/AI areas I work on:

• Pricing engine — a model estimating collaboration price ranges based on market data and creator profiles, combining several independent signals into a single output with communicated uncertainty (a range, not a single point estimate).
• Profile scoring and classification — multi-signal, interpretable classification of account types and content categories; a deliberate choice of an explainable model over a black box.
• Brand↔Creator matchmaking — assessing profile alignment with campaign briefs, supported by a language model with structured prompting and response validation.
• Semantic search — indexing profiles in a vector database for topical similarity to complement traditional filters.
• Campaign selection optimization — an algorithm that selects a combination of creators to maximize campaign impact within a given budget.

ML engineering practices I value:

• Decision logic versioning — to ensure historical results can be reproduced and new approaches can be safely tested alongside established ones.
• Characterization tests for computational logic — allowing for safe refactoring and model tuning without manual regression testing.
• Clean separation of domain rules from infrastructure — ensuring models are testable in isolation.