Google AI Studio works well for querying books, but over long conversations, it loops on the same info it retained in memory and loses access to the rest of the content — a classic RAG limitation. I'm looking for a turnkey RAG solution (not built from scratch) capable of searching, analyzing, and digesting the entirety of the sources with precision and nuance throughout the conversation, to then connect to Gemini as the LLM response generator. What existing RAG tool would you recommend for this?

Watching an idea I started with gradually take shape through building, experimenting, making mistakes, and improving has been an incredible learning experience. I’m especially grateful to the Decoding Data Science (DDS) mentors and @Mohammad Ahmad for the thoughtful feedback and suggestions during Tuesday’s demo session this week. Their perspectives are helping me make Kindred better as I move toward the finish line. 🙏 Just a couple more days to go, and I’m genuinely excited to see what the final product becomes! ❤️ #Kindred #DecodingDataScience #AgenticAI #BuildInPublic #LearningByDoing #AIInnovation

Watching Kindred evolve from an idea into something tangible has been an incredible journey of building, learning, feedback, and continuous improvement. I’m truly grateful to Decoding Data Science (DDS), the mentors, @Mohammad Ahmad and this amazing community for the guidance, encouragement, and opportunity to keep pushing ourselves. Wishing every fellow contestant the very best as we head into the final day—regardless of the outcome, everyone who dared to build, learn, and finish this journey is already a winner! #Kindred #DecodingDataScience #AgenticAI #AIInnovation #BuildInPublic #LearningByDoing

🚀 What started as an idea became Kindred, but the real reward has been the journey of building, experimenting, learning, receiving feedback, and pushing beyond my comfort zone. I’m deeply grateful to Decoding Data Science (DDS), @Mohammad Ahmad , the mentors, and this amazing community for making this experience possible. 🙏 To every fellow contestant: I wish you all the very best—and regardless of the final results, every one of us who showed up, took the challenge, built something, and made it to the finish line is already a WINNER! 🏆 #Kindred #AgenticAI #DecodingDataScience #AIChallenge #BuildInPublic #LearningByDoing #AIInnovation

Another productive day of refining the concept, strengthening the user experience, and bringing Kindred one step closer to becoming a meaningful, human-centered solution. 💡 Each day of the Decoding Data Science Agentic AI Application Challenge continues to reinforce the power of learning by building, experimenting, and improving. Excited about the progress so far—and even more excited for what comes next! 🙌 #Kindred #AgenticAI #DecodingDataScience #AIInnovation #NoCode #BuildInPublic