What I'm Reading

By

Hey,

  • This page is my personal archive of research papers I've read and found valuable.
  • Each entry includes the paper title, a link to the PDF/arXiv, and brief notes on why it mattered to me.
  • It's a living reading list – updated whenever I finish something noteworthy.
  • Think of it as a window into my research interests and learning journey.

    • 2025

    1. CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Selection in Text-to-SQL
      1. Many SQL drafts - AI makes lots of different SQL queries for one question using smart tricks.
      2. Pairwise pick - Another AI compares them two-by-two and chooses the best one (more accurate).
      3. Auto-fix errors - Grabs only needed data and fixes buggy queries using database feedback.
    2. Context Engineering: Sessions, Memory
      1. Context Engineering — Goes beyond basic prompts: smartly builds the full "world" an AI sees each turn (instructions, history, tools, memories) to make it act intelligently.
      2. Sessions — Short-term "desk" for one chat: holds active conversation, recent notes, and working state so the AI doesn't forget mid-talk.
      3. Memory — Long-term storage across chats: extracts key facts/preferences from sessions, saves them persistently, and loads them back for personalized, ongoing help.

    2024

    1. RAG-Fusion: a New Take on Retrieval-Augmented Generation
      1. RAG-Fusion enhances standard RAG by generating multiple related queries from the original one, then fusing retrieval results using Reciprocal Rank Fusion (RRF).
      2. Core advantage: Answers are more accurate and comprehensive because the model sees the question from several angles and prioritises consistently relevant documents.
      3. Main trade-off: ~1.8× slower than basic RAG due to extra LLM call and larger context.
    2. A Survey on LLM-as-a-Judge
      1. LLM-as-a-Judge means using a large language model (like ChatGPT) as an automatic judge to score or compare answers instead of humans.
      2. The paper organizes everything clearly: how these AI judges work, how to use them better (e.g., ask for pairwise comparisons or vote with several models), and how to fix their common mistakes (like favoring longer answers).

    Comments

    Post a Comment

    Popular posts from this blog

    Deploying AI Agents in Production Using Open Source Architecture

    By
    Image
    AI agents are becoming core parts of modern applications, but deploying them in production at scale is challenging. The diversity of agent frameworks, unpredictable latency, and the need for streaming responses make it hard to use simple REST-only patterns. Most proof-of-concept deployments break down when real workloads, traffic, and observability requirements hit. This blog post explains how to design a production-ready, open-source architecture for AI agents using FastAPI, Celery, Redis, Kubernetes, KEDA, Prometheus, Grafana, LangFuse, and LangGraph . Why Deploying AI Agents Is Hard Agent frameworks differ widely in required infrastructure. Latency varies from milliseconds to minutes depending on workflow complexity. Real-time streaming is needed for modern AI UX. REST-only patterns can’t handle long execution, retries, or async scheduling. Scaling compute-heavy agents is fundamentally different from scaling API servers. Why the Traditional REST Architecture Fail...
    Read more

    Welcome to Pods and Prompts

    By
    Image
    Hi everyone ðŸ‘‹, Welcome to my blog,  Pods and Prompts ! This is the space where I’ll be sharing my thoughts and experiences on Generative AI, Large Language Models (LLMs), and the infrastructure that powers AI applications . Everything I write here reflects my personal views  and is in no way related to  my past or present employers. A little about me: I’m Nithin Anil , an AI/ML engineer with over 13 years in the software industry . Over the years, I’ve worked with AWS, Azure, and GCP , built and managed large-scale distributed big data  systems , and designed microservice-based architectures . For the past couple of years, my focus has been on Generative AI applications,  and one of my strengths is managing infrastructure for highly available, low-latency AI systems . I’m also deeply experienced in Kubernetes and Infrastructure as Code . System design for millisecond or even microsecond latency is one of my favourite challenges. Beyond work, I’m pass...
    Read more

    Semantic caching for LLM Applications and AI Agents

    By
    Image
    Caching is one of the easiest ways to speed up applications and control cost. But LLM based systems don’t work well with traditional caching because users phrase the same idea in many different ways. This means most queries turn into cache misses. Still, caching is important. LLM agents can take time to run, and inference is expensive. A smarter caching method is needed. Why Traditional Caching Fails Works on exact matching. Natural language rarely matches exactly. Same intent written differently becomes a cache miss. Result: almost no benefit for LLM workloads. What Semantic caching Does Focuses on meaning instead of exact text. Convert each query into an embedding. Store embeddings in a vector database like Redis, Qdrant, or Milvus. Add TTL to control freshness. For each new query: Convert to embedding. Run similarity search. If similar enough, return cached output. Otherwise, run the agent. Redis Code Sample for Semantic caching Belo...
    Read more