◢ daily log

Build log

One post a day on building with AI: tools, agents, RAG, automation and what it takes to ship.

21 entries
  1. LOG_021 3 min read

    Architecting Persistent Memory for AI Agents with Supabase and Vector Embeddings

    Learn how to build persistent memory for your AI agents using Supabase and vector embeddings to overcome context window limitations.

    Read entry
  2. LOG_020 5 min read

    Integrating LLMs for Code Generation: Accelerating Your Workflow

    Learn how to integrate LLM-powered code generation into your existing development workflow to automate repetitive tasks and speed up common coding.

    Read entry
  3. LOG_019 4 min read

    Designing a Flexible and Scalable API for LLM Features

    Learn how to build a robust API for your LLM-powered features, focusing on flexibility, scalability, and maintainability as models evolve.

    Read entry
  4. LOG_018 6 min read

    Implement a Custom RAG Reranker for Better Relevance and Cost Efficiency

    Learn to implement a custom RAG reranking strategy using specialized models to improve relevance without increasing overall inference costs.

    Read entry
  5. LOG_017 4 min read

    Stream LLM Responses Without Breaking Server Logic

    Learn how to implement streaming UI updates from LLM responses while ensuring your critical server-side logic runs correctly on the full output.

    Read entry
  6. LOG_016 5 min read

    A/B Testing LLM Prompts in Production: A Must-Have System

    Learn why A/B testing LLM prompt variations is critical for production features and how to build a robust system to manage and evaluate them.

    Read entry
  7. LOG_015 5 min read

    Choosing the Right Embedding Model for Your RAG Application

    Learn how to select the best embedding model for your RAG application by analyzing your specific data characteristics and use cases.

    Read entry
  8. LOG_014 5 min read

    Evaluating RAG Retrieval with Human Feedback

    Learn how to implement human-in-the-loop feedback to accurately evaluate and improve retrieval quality in your RAG systems.

    Read entry
  9. LOG_013 5 min read

    Structuring and Embedding Multi-Modal Data for Advanced RAG

    Learn how to prepare and embed diverse data types like images, audio, and text to build more powerful RAG applications.

    Read entry
  10. LOG_012 6 min read

    Implement Robust RAG Data Pipelines with n8n

    Automate data ingestion and indexing for your RAG applications using n8n, ensuring data freshness and pipeline reliability.

    Read entry
  11. LOG_011 5 min read

    Build Reliable AI Agents with Self-Correcting Prompt Engineering

    Improve AI agent reliability by implementing self-correcting prompt engineering workflows to handle errors and refine outputs.

    Read entry
  12. LOG_010 5 min read

    Refining RAG with Supabase Vector Search Filtering

    Improve RAG relevance and reduce LLM costs by implementing vector search filtering with Supabase and pg_vector for precise document retrieval.

    Read entry
  13. LOG_009 5 min read

    Designing Effective Tool Schemas for AI Agents

    Learn how to design robust tool schemas for AI agents to prevent invalid function calls and improve reliability in your AI applications.

    Read entry
  14. LOG_008 4 min read

    Optimizing RAG Context Windows for Relevant Retrieval

    Improve RAG performance by intelligently chunking and embedding unstructured data to ensure only the most relevant information reaches your LLM.

    Read entry
  15. LOG_007 4 min read

    Building Efficient AI Agent APIs with TypeScript and Vercel Edge Functions

    Learn how to build low-latency AI agent APIs using TypeScript and Vercel Edge Functions for better performance and developer experience.

    Read entry
  16. LOG_006 6 min read

    Cost-Effective LLM Guardrails: Prompt Engineering and API Design

    Implement essential LLM guardrails cost-effectively using smart prompt engineering and robust API design, preventing misuse and controlling spend.

    Read entry
  17. LOG_005 5 min read

    Efficient Context Windows for AI Agents: Prompt Engineering Techniques

    Learn how to optimize context window usage for complex AI agent tasks, reducing costs and improving performance with practical prompt engineering

    Read entry
  18. LOG_004 4 min read

    Automating CI/CD with Generative AI: Code Reviews and Testing

    Integrate generative AI into your CI/CD pipeline to automate code reviews and enhance test generation, streamlining development workflows.

    Read entry
  19. LOG_003 6 min read

    Building a Production-Ready RAG System with Supabase and TypeScript

    Learn how to build a robust Retrieval-Augmented Generation (RAG) system using Supabase for vector storage and TypeScript for application logic.

    Read entry
  20. LOG_002 5 min read

    API vs. Local LLMs: Choosing for Feature Development

    Deciding between API-based and local LLMs for product features involves balancing latency, cost, and control to ship effective AI-powered applications.

    Read entry
  21. LOG_001 5 min read

    Debugging LLM Tool Use: Tracing Agent Decisions

    Learn to debug large language model agent tool use by tracing their internal decisions, identifying issues in tool selection or argument formation.

    Read entry