LangTalks
פודקסט טכני בעברית על בניית סוכנים ואפליקציות עם מודלי שפה. המנחים דנים בנושאים חמים בתחום הבינה המלאכותית והשפות הטבעיות. הפודקסט כולל ניוזלטר דו-שבועי וקהילת ווטסאפ לדיונים ושאלות.
פרקים
2 - How LLMs are developed
The evolution of Language Models and OpenAI's training techniques.
3 - Intro to Agents
Theory and benchmarks for zero shot, few shot, chain of thoughts, and tree of thoughts
4 - Intro to LangChain
Overviewing the main concepts of LangChain framework and how to get started
5 - How to use Embeddings
Explore the theory of embeddings and their practical applications
1 - Intro to LangTalks
Introduction about Lee & Gal and what LangTalks is about
6 - Security and Privacy | Itamar Golan (Prompt Security)
Security, privacy, compliance & safety of LLM-based applications
7 - Agents in LangChain | Eden Marco
From an agent POC with LangChain to production challenges
8 - Vector Databases | Uri Goren (Argmax)
Why you might need a VectorDB, which features to consider when choosing a provider, and use cases examples
9 - Popular tools for LLM-app devs
Save time building applications with LLMs by utilizing tools like LangChain, Llama index, Guidance, GPTCache and many more
10 - Summarizing mental health sessions | Amit Spinrad (Eleos Health)
Drill down on a summarization task use case
11 - Open source models | Matan Kleyman
why should you care about open source models, when and how to finetune, inference best practices, deployment & serving tools, models overview
12 - Fine-tuning | Mike Erlihson
Finetuning techniques (SFT, RLHF...), pros & cons, tools (technical episode)
13 - Production environment | Idan Tovi (Vee)
Full stack LLM app development, Typescript vs Python
14 - Prompt engineering
Advanced prompt engineering techniques: skeleton of thoughts, directional stimulus prompting, graph of thoughts, augmentations
15 - Initiatives and volunteers matching
Find the most suitable project for volunteer support "Iron swords" war efforts
16 - Multi modal
multimodality is a type of model that can analyze multiple data types like language, images, voice, etc.
17 - Evaluation | Philip Tannor (DeepChecks)
Evaluating LLMs and AI pipeline in dev and production environments. How to work with datasets
18 - OpenAI Dev Day
OpenAI DevDay announcements recap
19 - RAG
Retrieval Augmented Generation process, Llama index vs LangChain, indexing
20 - RAG Evaluation | Jonathan Yarkoni
ragas framework for evaluating unstructured retrievals and generations
21 - Knowledge Graph | Jesús Barrasa (Neo4j, ENG)
Indexing knowledgebases with KG for RAG applications
22 - LangChain v0.1
LCEL, LangGraph, LangSmith
23 - LLM theory | Mike Erlihson
What are transformers, why it is so expensive to train a Transformer-based model and what is the architecture of the future LLMs
24 - MVP best practices
how to define effective requirements for llm apps, and first steps after going to production
25 - Reranking
exploring the reranker component in the 2 stage retrieval system
26 - 12 RAG Pain Points and Proposed Solutions
The 12 most common challenges around building an effective RAG pipeline and the best practices for solutions
27 - Latency optimization | Dor Ringel
how to solve latency issues with minimal compromise on quality and cost
28 - Advanced agents
examples and practices of advanced agents, and use of LangGraph for effective tool usage by agents
29 - AI PM: trust & evaluation | Shira Weinberg Harel
build trust with users and stakeholders - transparency and explainability, reliability and consistency (how to fail safely), automation vs user control.
how to evaluate with uncertainty, prompt engineering, effective bug reports
30 - Open Source Security | Avi Lumelsky (Oligo)
everything about adopting open source code and models for your llm app
31- Multi agents | Assaf Elovic ("GPT Researcher")
Review of why and how to build a multi-agent system. Assaf is Head of R&D @ Wix and GPT-researcher open source creator.
32 - AI PM: No-code Tools | Galit Galperin
Utilize no-code tools for fast POCs
33 - LangGraph | Eden Marco
Getting started with LangGraph, the most popular multi agent framework
34 - Legal documents | Yaron Vazana (Darrow)
Build LLM-app with legal documents that have many references and domain verbiage
35 - LLM app dev guide | Almog Baku
end to end guide on how to get started and deploy to production your llm app
36 - MVP with open-source LLM | Yael Daihes (Anecdotes)
If you need to build an llm-app that uses an open-source llm, this episode is for you. Intermediate level
37 - Upgrading LLM | Dolev Pomeranz (Gong)
How to migrate the llm provider in your AI app safely
38 - OpenAI Realtime API
Technical review of the new Realtime API introduced on the dev day
39 - Code generation agents | Ziv Bakhajian
Cursor, RepoAgent, Aider, and some more coding agents - overview and interesting architectural concepts
40 - Advanced RAG | Nir Diamant
advanced RAG techniques like RAPTOR and using Clues
41 - GraphRAG
Deep dive on why GraphRAG approach was created and how it works
42 - Text to SQL
how to build a text to sql agent
43 - Building with GenAI-powered tools | Idan Benaun
this episode has 2 aspects - best practices to start a new product in 2025 and what tools can accelerate your process
44 - MCP intro
all you need to know about MCP and how to get started as a developer
45 - AI PM: Spec for copilot feature
Example of product specification required from an AI PM for a new feature
46 - Building coding multi-agent (LangTalks '25 conf)
Building SWE multi-agent with LangGraph: Zero to Hero | Lee & Gal (LangTalks GenAI 2025 Conference)
Lecture video:
https://www.youtube.com/watch?v=KroPK5DygWw
Repo:
https://github.com/langtalks/swe-agent
47 - Tech stack for Agents
Python or JS? LangChain or Vanilla? In this episode we gave guidelines on important things to notice when picking a tech stack for new project or a company's platform
48 - Vibe coding | Sahar Carmel
Tips for AI-driven coding
49 - Documents analysis
How to handle long docs like PDF and Docx effectively with LLMs
50 - A2A protocol
Is this Agent2Agent new protocol by Google going to be the next MCP?
51 - Our AI-driven coding workflow | Oren Melamed
Sharing our workflow with Cursor, Claude Code, GitHub Copilot and others. Listed the useful MCPs and tools, how to measure efficiency gains, and our vision for the future of coding
52 - Agents at scale | Adam Cohen Hillel
The story of building Suna - the open source alternative of Manus the generalist autonomous agent
https://github.com/kortix-ai/suna
53 - Search & Retrieval | Guy Itach (Fiverr)
techniques to combine classic search with agentic retrieval
54 - Permissions | Or Weis (Permit.io)
Enabling secured flexibility to your agents and MCP servers
55 - Context Engineering
The state of best practices for context engineering, focused on AI-coding as a leading use case
56 - n8n | Shay Shitrit
Getting started with no-code agentic automations, popular use cases and tips
57 - Memory | Itamar Friedman (Qodo)
Short and long term memory mechanisms for agents
58 - AWS re:Invent 2026 predictions | Shuki Cohen (AI21)
Recap of the major announcements and innovations from the 2025 re:Invent conference, along with our predictions for where things are heading in 2026
59 - SLMs | Dr Elad Granot (Trullion)
the production experience and considerations for fine-tuning a small language model for narrow tasks
60 - Brain Memory | Dr. Meytar Zemer
The Neuroscience behind the human brain and memories
61 - Voice Agents | Shay Davidson (Lemonade)
Building and testing voice agents with realtime API or chained approach (TTS->LLM->STT)
62 - AI R&D Rollout | Iko Azoulay (Salt Security)
How AI is used across the SDLC in a real R&D org - requirements, tech design, validation, testing, and code review. What worked, what didn’t, and where teams get stuck.
63 - WAKE UP
Singularity and AGI are around the corner.
64 - AI coding metrics | Liad Elidan (Milestone)
Measure the actual impact of adopting AI across the R&D
65 - AI SRE | Asaf Savich (Komodor)
Site Reliability Engineer agent auto fixes production issues
66 - Scaling LLMOps | Avi Lumelsky (Oligo)
Two real-world use cases are covered: CVE Enrichment for open-source vulnerability detection and real-time Malicious Command Detection, along with practical tips on cost optimization, prompt caching, and open-source vs. commercial model trade-offs.
67 - FinOps for AI | Yizhar Gilboa (Finout)
A practical look at managing AI budgets across coding agents, customer-facing models, and production inference
68 - AI-SDLC | Yonatan Maor (Clears.ai)
A practical, stage-by-stage look at running your entire development lifecycle with AI agents: from PRD to coding to QA.
69 - Marketing for Agents | Nevo David (Postiz)
How to market and sell to AI agents — discovery channels, CLI vs MCP trade-offs, authentication patterns, and the emerging agent-driven economy.
70 - Our Claude Code Tips
A behind-the-scenes look at real agentic-coding workflows: visual planning, cloud-first agents, ephemeral PR environments, custom skills, and secrets management.
71 - Claw Architectures | Gavriel Cohen (NanoClaw)
A deep dive into the architecture and security model behind autonomous agent harnesses - isolation, credential gateways, and skill-based extensibility.
