AI Engineering Intern

What I worked on

LINQIA is an influencer-marketing platform connecting brands with creators at scale. I joined the AI team during a short, intense internship to ship two pieces of capability.

Creator–brand matching

• Analysed millions of creator profiles with NLP + computer vision techniques
• Matched content styles to brand-campaign requirements using a weighted scoring model
• Built an LLM-driven brief generator: takes brand guidelines + historical campaign performance and produces personalised creator briefs
• Auto-validated creator submissions against the brief requirements

VideoSpec — multimodal video analysis

A platform for ingesting and structurally analysing creator video content across TikTok, Instagram, and direct MP4:

• Custom URL extraction handling each platform's quirks (auth, anti-bot, format variation)
• Video embedding for semantic search and similarity
• 15+ classification parameters per video (style, pacing, brand-safety, content category, hook quality)
• Asynchronous processing pipeline using AWS Nova + Gemini Vision API for parallel multimodal inference
• Streamlit front-end for the analyst team, OAuth2-authenticated Google Sheets integration for non-technical stakeholders

Highlights

• Shipped end-to-end in roughly 4 weeks
• Helped the team move from manual creator vetting to automated, multimodal scoring
• Worked from LINQIA's San Francisco office during the trip; remainder remote