Share

The Brainforge Way: November's update on events and the latest news in AI

Langfuse AI Gateway; what your sales calls already know; analytics drift fix

Brainforge

November edition

The Brainforge Way

A quick hello

Hey there,

Whether you’re a partner, a client, a friend, or someone we’ve been hoping to meet, here’s what we’ve been building, hosting, and learning lately.

In this issue

•

Click.Forge.Fuse recap: What a room of people running agents in production wanted to see.

•

From the blog: Langfuse AI Gateway first, then Transcript Intelligence and product analytics drift.

•

A client, in their words: ABC Home & Commercial on how we improved their support team

•

Two ways we can help: AI observability and product analytics, if these problems sound familiar.

•

Brainforge Banter: Two questions to take into your next team meeting


01 / What we are seeing

Event recap: Click.Forge.Fuse in Austin

Click.Forge.Fuse event graphic
ClickHouse
Google
Langfuse

Your agent ran last night. Can you say what it did, why it failed, and what it cost? That was the question in the room on October 21, when we hosted Click.Forge.Fuse at The Pershing in Austin with ClickHouse, Google, and Langfuse.

The agenda stayed on that problem. A customer recapped trade surveillance at petabyte scale on ClickHouse, attendees saw live demos and spoke with engineers who run these systems, and everyone went home with a custom gift. Missed it? We’re planning a November workshop. Details are in “Where to find us” below.


Pendo

Product reviews that start in the right place “Where do we even look?”

On October 13, Pendo posted about product reviews that open with that question. Their answer: App health is now generally available, along with AI adoption views for agents and Add to Dashboard.

That puts the first view in one place. The next step is getting that signal into the warehouse, so product, data, and go-to-market teams share one definition of “working.”

Here’s what it looks like when they don’t. A veterinary compliance platform was counting the same journeys three different ways in Pendo. Our audit found 68 track events, 76 pages with overlapping rules, and 256 features in the taxonomy. It ended with one event plan tied to four real milestones: sign-up, license check, signed certificate, and billing.

Read Pendo’s post

·

Read the veterinary compliance case study

New on the blog

Langfuse

Change two lines. Every model call becomes a trace.

Langfuse put its AI Gateway into public beta. Point your app or coding agent at Langfuse instead of OpenAI or Anthropic, keep the request body the same, and change the base URL and API key. Shared provider keys go away, and every call that used to be invisible lands as a trace you can open.

Read the Langfuse AI Gateway post

Calls are a revenue system, if you can read them

Most teams already record calls. But the record often just says “called,” while the conversation had the intent, the objection, and whether anyone booked a next step.

Our recent post shows how to turn that audio into a record marketing, operations, and sales can all use: classify the lead, pull the qualification signals, mark the outcome, and send it to someone who will act on it. Start small. A few labels like qualified lead, appointment booked, and follow-up required will do more in week one than a model that scores 40 emotions.

Read Transcript Intelligence

Product analytics drifts, then the review argues about the chart

You’ve probably seen the signs. The same behavior shows up under more than one event name. Properties go missing or change type. Old events linger in dashboards. And “activation,” “retention,” and “conversion” mean different calculations to different teams.

Our post is a 2-week reset for when events, properties, dashboards, and definitions stop agreeing.

Read the product analytics drift post

A client, in their words

ABC Home & Commercial

Read the case study


Two offers

If the Austin night, the product review, or the posts sound like your week, these are the two places to start.

•

AI observability. Traces, evals, and monitoring for LLM apps and agents, so you can see what an agent did, why it failed, and what it cost. AI observability

•

Product analytics. Event taxonomy, instrumentation, and activation, so a product review starts from one definition of working. Product analytics

03 / See it in action

Check out our demos page!


04 / Where to find us

Recap and upcoming

Vixul

Recap: AI Native Services Summit

Nov. 16, The Pershing

Austin

Read the recap

Langfuse

Langfuse Hands-On Workshop

Nov. 19, 9–11 a.m. CT

WeWork Domain + Zoom

Austin

RSVP

AWS

AWS re:Invent

Nov 30-Dec 4

Las Vegas

Say hello if you’ll be there.

Online

LinkedIn:

Website:

Blog:

Put this into practice

Bring one of these ideas to your own data.

We scope a single use case, build it, and show you it working in weeks. Here's where to start.

Put this into practice

Bring one of these ideas to your own data

We scope a single use case and show you how it works, in weeks. Here is the shape of that work for a team like yours.

Book a scoping callAI →See pricing →
Shipping

Clean Data for Smarter Shipping Ops

How Brainforge helped Stella Source unify quoting, CRM, and operations data into a governed analytics hub that cut quote turnaround time 60% and improved cost-estimation accuracy.

  • 60%reduction in quote turnaround time
  • 95%cost estimation accuracy achieved
  • <24 hrstime to actionable insights

Put the idea to work

Turn what you learned into a practical next step.

We can help you identify the right starting point, scope the work, and ship something useful without committing to a large transformation first.

Press Esc to close