NeoSky AI @NeoskyAI

Jensen closed his GTC Berlin 2026 keynote with this Hermes Agent demo as the finale. He introduced it by saying the way to build an agent is to ask an agent to build it for you, so he installed and set up a customized Hermes in one prompt. Then he used Hermes, running on @nvidia RTX Spark, to create a simulated training environment for the @huggingface Microduck. Hermes planned an obstacle course with Nemotron and had GPT-6 Astra build it using FreeCAD, Omniverse libraries, Blender, Newton, and Isaac Lab. Hermes trained thousands of simulated ducks on that course until the validated policy was ready to run on the real robot in the physical world. At Nous, we believe open source is a practical tool as much as a philosophical ideal. It gives users the ability to build powerful, complex workflows chaining numerous local+cloud models, devices, and tools themselves instead of waiting for a company to enable them.

In healthcare intake, a good summary attached to the wrong patient record is still a failure. Separate collecting details from matching a record. Check identifiers and conflicts before write-back; if the match is unresolved, don’t guess.

@NeoskyAI's FalconBuilder.dev platform and usage updates from May 22, 2026 to September 24, 2026 The headlines: - Executions: 561,607 in production, up 20.7x from 27,128. Daily volume averaged about 4,276, up from roughly 349 in the spring. - Adoption: users grew 83% to 77, workflows grew 4.9x to 805, and agents reached 106. The average user now runs 10.5 workflows, up from 3.9. - Product: 519 commits shipped. Highlights include a self-hosted Community edition, the Falcon Builder MCP server, workflow versioning with rollback, and a verification and approval layer for regulated operations.

Falcon Builder now has an AI Decision node using Jev AI to allow quick structured decisions directly in workflows. See the doc: falconbuilder.dev/docs/integrati…

I just published Self-Host Falcon Builder on Your Infrastructure medium.com/neosky-ai/self…

BAA and zero-data-retention are not compliance badges. They’re architecture choices. Regulated AI agents need trust boundaries before the model call: minimize data, pick the right path, keep audit evidence, and avoid sending sensitive context where it does not belong.

We just launched a self-hosted version of FalconBuilder.dev that you can now run on your own machine. Run: git clone github.com/NeoSky-AI/falc… cd falcon-builder-self-hosted ./setup.sh # asks for your public URL, generates every secret into .env docker compose up -d

I just launched a self-hosted version of FalconBuilder.dev that you can now run on your own machine. Run: git clone github.com/NeoSky-AI/falc… cd falcon-builder-self-hosted ./setup.sh # asks for your public URL, generates every secret into .env docker compose up -d

The most important skills for using AI coding agents effectively. Presenting the AI Engineering Skills Map for using coding agents. x.com/i/article/2095…

Before an AI agent sends a text, email, or call, it needs a communication policy layer: consent, opt-outs, channel preference, quiet hours, workflow stage, delivery failure handling, and record updates. The agent reasons. The policy layer enforces the rules.

no self-hosting, no ceiling on what you pay. a client at 50k executions a month is burning hosted-plan money for infra that fits on a $20 VPS. volume triples and you are just hostage to someone's pricing page

Enterprise AI often starts too late. Before an agent decides or acts, messy intake has to become a clean work packet: request, context, missing data, sensitive fields, source record, and retained evidence. Reliability starts there. #EnterpriseAI

templates are the difference between a prospect and a signed contract. you walk into a client call with a lead-intake workflow already connected to their CRM, show them their inbox sorted every morning, and they stop imagining and start saying yes

@Relay Is Shutting Down: Dates, Export Guide, and Where to Go Next falconbuilder.dev/blog/relay-app…

We're hiring @NousResearch The work is hard and the stakes are high. Help us build a different future. Software Engineer Software Engineer, Product Software Engineer, Hermes Cloud UI/UX Designer and more: nousresearch.com/careers

If any teams/companies/individuals are being affected by @relay shutting down, we'd be happy to help you migrate to @FalconBuilderHQ. We'll help you out. Take a look at our migration process: falconbuilder.dev/relay

Not every step in an AI agent workflow should be AI. Use deterministic rules for known business logic, AI for messy context and judgment, then validate before action. Regulated operations need hybrid systems, not model-only workflows. #EnterpriseAI

Introducing Bot Mode for Hermes Desktop. Your agent profiles become a series of named Bots. Each Bot has its own role, model, memory, skills and profile picture; Bots can use any model and even communicate with each other. Build a specialist Bot once to use it forever.

A game changer for those that are building agents for clients that have similar functionality and needs.

Mark E. Chiles @markechiles

2 months ago

Did I mention we added Skills to @FalconBuilderHQ so you build/import skills to be used across multiple workflows and agents? See here 👉 falconbuilder.dev/docs/skills

0 0 4 66 0

Are you looking to get your data in order for AI? Meet Data Stew - datastew.ai