AI by Hand ✍️
“AI by Hand ✍️” delves into the intricacies of building AI applications with a focus on memory, context, and persistence. The document explores the importance of short-term memory in AI applications, specifically in maintaining conversation history for seamless interactions. It highlights the role of tools like Langchain in enabling AI agents to continue conversations from where they left off, enhancing user experience. Additionally, the document discusses the significance of leveraging databases like PostgreSQL for storing conversation details and managing context effectively. Overall, “AI by Hand ✍️” provides insights into the practical implementation of AI technologies for enhanced user engagement and functionality.
Kimi 3 (Aug 6, 2026)
AI by Hand ✍️ Seminars
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Context Problems
Agentic AI by hand ✍️
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Cost Forecast
Budgeting a Long Session
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Quadratic Cost
Adding Up the Turns
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Retrieval Cost
Retrieved Chunks Are Input
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Stateless API
Resending the Context
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Verbatim Tail
Keeping the recent rounds whole
📚 Read more at AI by Hand ✍️🔎 Find similar documents
The Compaction Threshold
Picking the number from a round
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Compacted Retrieval
A summary instead of raw chunks
📚 Read more at AI by Hand ✍️🔎 Find similar documents
The Compaction Bill
What a summary call costs
📚 Read more at AI by Hand ✍️🔎 Find similar documents
Compaction
Summarizing history to make room
📚 Read more at AI by Hand ✍️🔎 Find similar documents