Currently Building

India's career market is broken on both sides. Candidates lack structured paths to break into product/AI roles, and recruiters spend weeks chasing the wrong shortlists. Cohort-based learning is too expensive for most Indian learners. AI matchmaking products are built for the US labor market, not Indian context, languages, or salary bands.

• ●Praxel Learn live at praxel.in — cohort-based AI/product learning for Indian learners.
• ●Praxy live at praxy.praxel.in — single product, both sides of the marketplace (candidates find recruiters, recruiters find candidates).
• ●Indic TTS/STT research powering Praxy's voice flows (see Research).

• →Scale Praxy's matching across 10+ Indian languages.
• →Open Praxel Learn cohorts for Q3 2026.

Scrapers break, rotate, and get blocked. Most teams either build brittle in-house infrastructure or pay for opaque scraping APIs. Neither survives the agentic era — agents need scrape primitives that adapt without a human in the loop.

• ●Thirdwatch.dev live — agentic scraping platform.
• ●Apify actor catalog at apify.com/thirdwatch covering recruiting, B2B SaaS reviews, e-commerce, real estate, and food delivery.
• ●MCP surface so agents can call Thirdwatch as a first-class tool.
• ●Docs and blog with category-specific guides for each vertical.

• →Open the agent-facing API for autonomous scraping pipelines.
• →Expand the actor library across more verticals.

Communication is the highest-leverage skill in every knowledge job, and the slowest to improve. Books and courses don't move the needle. Real practice with real feedback does — but most people never get it.

• ●Commy.app live — practice conversations with AI feedback.
• ●Scenario library across difficult conversations, interviews, and presentations.

• →Add longitudinal tracking so users see growth over months, not minutes.
• →Voice-first practice flows.