SLMs vs. LLMs: Why Smaller AI Models Win in Business Once you stop asking AI models to understand everything, they get much better at understanding what matters.

Once you stop asking AI models to understand everything, they get much better at understanding what matters.

By @seanfalconer.bsky.social, thanks to @confluent.io

Real-Time Anomaly Detection with Apache Flink Catch critical issues before they impact revenue — learn how to detect real-time payment failures using Apache Flink and Confluent Cloud.

Real-Time Anomaly Detection with Apache Flink - @seanfalconer.bsky.social

seanfalconer.medium.com/real-time-an...

The future of the agentic enterprise runs on real-time data.
Join Salesforce & Confluent live at #Current25 as Gunther Hagleitner and @seanfalconer.bsky.social share how real-time context powers the next generation of agentic AI systems. Stream it free on 10/29→ https://cnfl.io/43zOrWP

AI that thinks fast needs data that moves fast.
At #Current25 NOLA, @anthropic.com Rachel Lo will join @seanfalconer.bsky.social on the mainstage to show how Claude + data streaming are changing what’s possible for intelligent systems. Don’t miss it -> https://cnfl.io/3WH6mag

SED News: Corporate Spies, Postgres, and the Weird Life of Devs Right Now - Software Engineering Daily Welcome back to SED News, a podcast series from Software Engineering Daily where hosts Gregor Vand and Sean Falconer break down the latest stories in software engineering, Silicon Valley, and wider te...

In this episode, Gregor Vand and @seanfalconer.bsky.social dive into the Deel and Rippling saga, Postgres bets by Databricks and Snowflake, this weird time for devs, and highlights from Hacker News.

softwareengineeringdaily.com/2025/06/17/s...

Crew AI with João Moura - Software Engineering Daily Agentic AI is seen as a key frontier in artificial intelligence, enabling systems to autonomously act, adapt in real-time, and solve complex, multi-step problems based on objectives and context. Unlik...

João Moura is the founder of CrewAI which is a platform for building and deploying automated workflows. He joins @seanfalconer.bsky.social to talk about the company’s rise in the agentic AI space and the future of automation.

softwareengineeringdaily.com/2025/06/03/c...

Real-Time Data You Can Prompt Using MCP YouTube video by Block Open Source

Watch the full episode to see how Edward Vaisman and @seanfalconer.bsky.social used Goose to explore Kafka data without writing a single line of code.

www.youtube.com/watch?v=ivlz...

Turing Award Special: A Conversation with Jack Dongarra - Software Engineering Daily Jack Dongarra is an American computer scientist who is celebrated for his pioneering contributions to numerical algorithms and high-performance computing. He developed essential software libraries lik...

Jack Dongarra is a Turing Award winner who is celebrated for his pioneering contributions to numerical algorithms and high-performance computing. He joins the podcast with @seanfalconer.bsky.social to talk about his life and career.

@acm.org

softwareengineeringdaily.com/2025/03/18/t...

NVIDIA RAPIDS and Open Source ML Acceleration with Chris Deotte and Jean-Francois Puget - Software Engineering Daily NVIDIA RAPIDS is an open-source suite of GPU-accelerated data science and AI libraries. It leverages CUDA and significantly enhances the performance of core Python frameworks including Polars, pandas,...

Chris Deotte is a Senior Data Scientist and
@jfpuget.bsky.social is the Director and a Distinguished Engineer at Nvidia. They join @seanfalconer.bsky.social to talk about NVIDIA RAPIDS and GPU-acceleration for data science tools.

softwareengineeringdaily.com/2025/03/04/n...

Multi-agent systems promise to solve complex problems, but only with the right foundation.

Event-driven design gives multi-agent systems a shared language, using events to decouple producers and consumers.

www.infoworld.com/article/3808...

Every System is a Log: Avoiding coordination in distributed applications Distributed coordination makes application complex and brittle. Because all systems eventually build on logs, we can use a shared-log approach to eliminate most coordination. We discuss this conceptua...

Distributed systems are complex, but what if most are just logs at their core?

"Every System is a Log" explores how embracing this idea reduces coordination and simplifies design.

Worth a read: restate.dev/blog/every-s...

A Practical Guide to Choosing Prompts, Workflows, or Agents Note: I recently contributed with my colleague Jack Vanlightly on an article exploring when it makes sense to use an AI agent. The topic…

AI agents are powerful, but they’re not always the answer. Sometimes a prompt or workflow is simpler, faster, and smarter.

Prompts: great for one-offs.
Workflows: solid for predictable tasks.
Agents: messy, adaptable, but tricky to scale.

Don’t over-engineer. More here: medium.com/@seanfalcone...

AI Agents in 2025 — Jack Vanlightly Two interesting blog posts about AI agents have caught my attention over the last few weeks. Anthropic wrote Building Effective Agents. Chip Huyen wrote Agents . Ethan Mollick has also...

The pace of AI advancement is not slowing down, and now AI agents are breaking onto the scene, leveraging ever more powerful models to interact with the real world. In this post, I cover some of the voices talking about AI agents and the challenges ahead.

jack-vanlightly.com/blog/2025/1/...

AI can't deliver value if your data is siloed, stale, and stuck in brittle pipelines.

Real-time streaming solves this by liberating data at the source. No batch delays. No reverse ETL headaches.

Check out my article for more on this.

medium.com/@seanfalcone...

Building a Research Assistant With Kafka and Flink Learn about the development of an AI-powered research assistant created to help prepare for podcast interviews.

AI is only as good as its data. I built a research assistant with Kafka + Flink to process real-time, in-flight data—no stale datasets, no rigid pipelines. Just clean architecture powering practical AI.

dzone.com/articles/bui...

It's built with:
◆ OpenAI GPT-4 model and Whisper
◆ LangChain for prompt templates and LLM API abstraction
◆ Next.js by Vercel
◆ Confluent Cloud for Kafka and Flink

I used an event-driven design to keep everything modular and scalable, de-coupling the user-facing application from the AI workflow. To do this, I used Confluent Cloud’s Data Streaming Platform, which supports Kafka, Flink, and AI models as first class citizens.

I don't always have time to craft a great post about a podcast I host. To make life easier, I built an AI-powered LinkedIn post generator.

Blog post: seanfalconer.medium.com/automating-p...

GitHub repo: github.com/thefalc/podc...

C# Compiler and Language Design at Microsoft with Jared Parsons YouTube video by Software Engineering Daily

C# Compiler and Language Design with Jared Parsons (Microsoft): youtu.be/IyMV01Y2PI8

Boston Dynamics with Matt Malchano: youtu.be/RwbvRbCE3aM

Linux on Apple Silicon with Alyssa Rosenzweig: youtu.be/ADMxNT8PmW8

TSMC and the Geopolitics of the Chip Industry with Tim Culpan: youtu.be/1hAqcFrhhAo

The Real Work of Data Engineering with Joe Reis YouTube video by Software Huddle

🎙️ This year, I hosted nearly 100 podcasts!

Here are some of my favorites👇

The Real Work of Data Engineering with @joereis.bsky.social: youtu.be/KwP4q00B6OI

One Year of ChatGPT with Christian Hubicki: youtu.be/cJP5341-GGw

Stop Treating Your LLM Like a Database A look at why the batch paradigm is a relic of the past, how it hinders AI apps and why the future of AI demands a real-time event-streaming platform.

While databases are introverts, holding information close until explicitly asked, LLMs are extroverts, designed to engage, synthesize and proactively contribute.

For more on this, check out my post on @thenewstack.io:

thenewstack.io/stop-treatin...