Anurag Tripathi

Full-time AI Product Manager, Senior Product Manager, Technical Product Manager / TPM, or Forward Deployed Product Manager roles at AI-first companies. Worldwide remote from India (async-first, EU/US overlap), available within 30 days. I bring shipped LLM products, an evaluation-led quality practice, and founder-level ownership from discovery to launch. At Xenriq AI Studio, I lead product strategy and delivery across four AI products used by 50,000+ people in 74 countries. My lead product, Promptive Sentry, is a prompt-engineering and context suite for Claude, ChatGPT, and Gemini power users — 20,000 users and a 4.6-star rating on the Chrome Web Store, grown with zero paid acquisition. I defined its core capabilities, including prompt enhancement and cross-platform context memory, and built trust into the product: local-first, no account required, no data selling. For Promptive Sentry for IDEs, I built a seven-layer context pipeline that runs under 800ms and cuts prompt iteration time by 60% (2,777+ downloads, 5.0 stars). I also built Piqque, a Windows ambient AI assistant with push-to-talk voice across Claude, Gemini, and GPT models. My AI product practice is evaluation-led: eval sets, failure modes, regression checks, guardrails, and human-in-the-loop design, with rollback plans before anything ships. Earlier, at Aithent Technologies, I partnered with product and design to launch an LLM-assisted help search that reduced support tickets by 12% and increased self-service usage by 10%, and I built the model evaluation framework behind it. At Xoriant, I owned delivery of a spending categorisation feature that reached 20% DAU within three months and became a top-3 engagement driver. I work across three modes: AI Product Manager, Technical PM (cross-functional program delivery, release and incident management, ex-Senior Software Engineer), and Forward Deployed PM (direct customer discovery and feedback loops with the developers and creators who use my products). Based in India, working worldwide remote — async-first, with EU and US timezone overlap.