Projects

At Pazago, I directed product and GTM for Provue AI. I built the underlying workflows, agents, and skills, eventually creating a system that let others build agents with any desired capability, and helped scale the product to 300k+ active users.

• Workflows & agentic systems: built the underlying workflows, agents, and skills, creating a system that let others build agents with any desired capability.
• Growth & GTM loops: structured high-performance organic loops that drove 5M+ views through strategic creator integrations and visual content cycles.
• Growth infrastructure: implemented precise event tracking and cohort funnels to cut onboarding drop-off.

Operating at the frontlines of product and GTM taught me to align user delight with distribution loops from day one. More at provue.ai.

Siftl is a save-anything app I built and shipped on my own. Share a link, a post, a screenshot, or a PDF from any app on your phone and it lands in one place, cleaned up and readable. Later, it comes back to you without you going looking for it.

• Share-sheet capture: an iOS share extension that saves articles, videos, podcasts, threads, and screenshots without leaving the app you are in.
• Automatic enrichment: titles, authors, artwork, and tags are pulled on save, so there is nothing to file, name, or sort.
• Resurfacing: saved items return through a home-screen widget and timed nudges, so the collection stays read rather than stored.

Built end to end, from the web app and enrichment pipeline to the native SwiftUI iPhone app. Free on iPhone at siftl.com and the App Store.

Acharya acts as a personal tutor for online video and audio. By ingesting YouTube lectures, podcasts, or local audio uploads, it constructs a fully interactive classroom.

• Content ingestion pipelines: handlers that scrape, transcribe, and index raw YouTube links or direct audio files.
• Interactive study loops: context-aware prompting that generates notes, flashcards, quizzes, and Q&A on demand.
• Minimalist interface: a clean, lightweight UI focused on video consumption alongside rapid Q&A.

Acharya lowers the barrier to active recall from long-form educational video. Try it at myacharya.vercel.app.

An interactive explainer synthesizing Ribbit Capital's June 2025 letter. It breaks down the taxonomy of stablecoin settlement pipelines, proprietary data-token factories, and context flywheels.

• Machine-legibility taxonomy: modeled Access, Memory, Expert, Context, Identity, Knowledge, and Asset tokens.
• Interactive data interfaces: reactive charts rendering inference-cost collapses (1,000x over three years) and settlement-volume thresholds.
• Gamified retention loops: a client-side quiz testing users on data moats and agentic resource allocation.

It maps the transition where vertical token networks dismantle traditional BPO and software ERPs. Explore it at thetokenrevolution.vercel.app.