The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.
Python agentic-memoryagentic-memory-systemagentsaiai-agentschatgptgenaillmlong-term-memorymemorymemory-managementpythonragstate-management
Add Unacademy AI demo (#1043)
Deshraj Yadava10823d309eb
diff --git a/examples/unacademy-ai/README.md b/examples/unacademy-ai/README.mdnew file mode 100644index 000000000..013f0322f--- /dev/null+++ b/examples/unacademy-ai/README.md@@ -0,0 +1,19 @@+## Unacademy UPSC AI++This directory contains the code used to implement [Unacademy UPSC AI](https://unacademy-ai.streamlit.app/) using Embedchain. It is built on 16K+ youtube videos and 800+ course pages from Unacademy website. You can find the full list of data sources [here](https://gist.github.com/deshraj/7714feadccca13cefe574951652fa9b2).++## Run locally++You can run Unacademy AI locally as a streamlit app using the following command:++```bash+export OPENAI_API_KEY=sk-xxx+pip install -r requirements.txt+streamlit run app.py+```++Note: Remember to set your `OPENAI_API_KEY`.++## Deploy to production++You can create your own Unacademy AI or similar RAG applications in production using one of the several deployment methods provided in [our docs](https://docs.embedchain.ai/get-started/deployment).diff --git a/examples/unacademy-ai/app.py b/examples/unacademy-ai/app.pynew file mode 100644index 000000000..5f0601672--- /dev/null+++ b/examples/unacademy-ai/app.py@@ -0,0 +1,97 @@+import queue++import streamlit as st+from embedchain import Pipeline as App+from embedchain.config import BaseLlmConfig+from embedchain.helpers.callbacks import StreamingStdOutCallbackHandlerYield, generate++[email protected]_resource+def unacademy_ai():+ app = App()+ return app+++app = unacademy_ai()++assistant_avatar_url = "https://cdn-images-1.medium.com/v2/resize:fit:1200/1*LdFNhpOe7uIn-bHK9VUinA.jpeg"++st.markdown(f"# <img src='{assistant_avatar_url}' width={35} /> Unacademy UPSC AI", unsafe_allow_html=True)++styled_caption = """+<p style="font-size: 17px; color: #aaa;">🚀 An <a href="https://github.com/embedchain/embedchain">Embedchain</a> app powered with Unacademy\'s UPSC data!</p>+"""+st.markdown(styled_caption, unsafe_allow_html=True)++with st.expander(":grey[Want to create your own Unacademy UPSC AI?]"):+ st.write(+ """+ ```bash+ pip install embedchain+ ```++ ```python+ from embedchain import Pipeline as App+ unacademy_ai_app = App()+ unacademy_ai_app.add("https://unacademy.com/content/upsc/study-material/plan-policy/atma-nirbhar-bharat-3-0/", data_type="web_page")+ unacademy_ai_app.chat("What is Atma Nirbhar 3.0?")+ ```++ For more information, checkout the [Embedchain docs](https://docs.embedchain.ai/get-started/quickstart).+ """+ )++if "messages" not in st.session_state:+ st.session_state.messages = [+ {+ "role": "assistant",+ "content": """Hi, I'm Unacademy UPSC AI bot, who can answer any questions related to UPSC preparation. Let me help you prepare better for UPSC.\n+Sample questions:+- What are the subjects in UPSC CSE?+- What is the CSE scholarship price amount?+- What are different indian calendar forms?+ """,+ }+ ]++for message in st.session_state.messages:+ role = message["role"]+ with st.chat_message(role, avatar=assistant_avatar_url if role == "assistant" else None):+ st.markdown(message["content"])++if prompt := st.chat_input("Ask me anything!"):+ with st.chat_message("user"):+ st.markdown(prompt)+ st.session_state.messages.append({"role": "user", "content": prompt})++ with st.chat_message("assistant", avatar=assistant_avatar_url):+ msg_placeholder = st.empty()+ msg_placeholder.markdown("Thinking...")+ full_response = ""++ q = queue.Queue()++ def app_response(result):+ llm_config = app.llm.config.as_dict()+ llm_config["callbacks"] = [StreamingStdOutCallbackHandlerYield(q=q)]+ config = BaseLlmConfig(**llm_config)+ answer, citations = app.chat(prompt, config=config, citations=True)+ result["answer"] = answer+ result["citations"] = citations++ results = {}++ for answer_chunk in generate(q):+ full_response += answer_chunk+ msg_placeholder.markdown(full_response)++ answer, citations = results["answer"], results["citations"]++ if citations:+ full_response += "\n\n**Sources**:\n"+ sources = list(set(map(lambda x: x[1], citations)))+ for i, source in enumerate(sources):+ full_response += f"{i+1}. {source}\n"++ msg_placeholder.markdown(full_response)+ st.session_state.messages.append({"role": "assistant", "content": full_response})diff --git a/examples/unacademy-ai/requirements.txt b/examples/unacademy-ai/requirements.txtnew file mode 100644index 000000000..bba32905a--- /dev/null+++ b/examples/unacademy-ai/requirements.txt@@ -0,0 +1,3 @@+embedchain+streamlit+pysqlite3-binary\ No newline at end of file