LangChain & Agents
LCEL chains, conversation memory, and tool-calling agent architectures
📝 LangChain Summarizer ⓘ
LCEL pipe chain: PromptTemplate → ChatOpenAI → StrOutputParser.
Summarize any passage into 2–3 concise sentences using a declarative LangChain Expression Language pipeline.
💬 Memory Chat ⓘ
Multi-turn conversation memory using ChatPromptTemplate with MessagesPlaceholder.
Introduce yourself or share context, then test recall across follow-up turns using client-persisted message history.
🔧 OpenAI Tool Agent ⓘ
Manual OpenAI Function Calling loop dispatching a local Python get_price(item) tool.
Ask prices for shoes ($50), hat ($20), bag ($35), shorts ($5), or pants ($10) from the mock store catalog.
🌤️ LangChain One-Tool Agent ⓘ
LangGraph create_react_agent equipped with a @tool get_weather(city) lookup.
Compare queries that trigger the weather tool (New York, London, Tokyo, Paris, Mumbai) against general knowledge answered directly.
🤖 LangChain Multi-Tool Agent ⓘ
Multi-tool ReAct agent orchestrating get_weather, add, multiply, and get_population.
Ask a compound question combining weather, arithmetic, and city populations in a single reasoning turn.