Anvesha
अन्वेषा means "The Explorer" in Sanskrit. A deep research API that searches 7 engines simultaneously and synthesizes AI answers from real page content.
What makes Anvesha different?
- Multi-engine search — Queries DuckDuckGo, Felo.ai, Bing, Brave, Mojeek, and Google Scholar, Startpage in parallel
- Real content extraction — Fetches actual page content, not just snippets
- AI synthesis — Generates answers grounded in verified sources
- Deep research mode — Auto-generates research questions and creates comprehensive reports
- Real-time progress — Watch each engine respond live as the search happens
Creator
I'm a 12th grade CS student at St. Xavier's Inter College, passionate about building AI-powered tools that solve real problems. I create YouTube tutorials on Python automation and AI assistants, and I'm constantly exploring how AI systems can better interact with users and their environments.
Anvesha was born from my frustration with AI hallucinations. I wanted a search tool that actually reads the pages it cites — not just guesses from snippets. This project combines my love for backend engineering, API design, and practical AI applications.
☕ Support This Project
If Anvesha has helped your research or projects, consider supporting!
Tech Stack
- Backend — Python, Flask
- Search — DuckDuckGo, Felo.ai, Bing, Brave, Mojeek, Google Scholar, Startpage
- AI — DeepInfra (Llama 3.1 70B)
- Frontend — Vanilla HTML/CSS/JS
- Streaming — Server-Sent Events (SSE)