All Posts

Why We Don't Want Leaky Models Here at Factal, we love data. We love machine learning. And, let’s be honest, we love seeing validation scores go up. There’s a distinct dopamine hit when you look at a metric dashboard and see accuracy hovering around 97%, 98%, or even the mythical 99%. For a brief, shining moment last week, we thought we were gods. Our latest batch of models - designed to automate some internal categorizations - weren't just performing well; they were operatin

Building an automated data analyst is a lot like being a digital detective. You want to give it a pile of data, walk away to grab a coffee, and come back to a whiteboard filled with "Aha!" moments. But as we learned recently, even the smartest AI can’t find a heartbeat in a mannequin. Here is the story of how Factal Signals  refused to lie to us, and why that makes it the most honest member of our team. What is Factal Signals, Anyway? Factal Signals Before we get into the "dr

Let’s talk about the specific kind of dread you feel when you open a new database schema for the first time. You know the feeling. You’ve been promised a "structured data warehouse." You have visions of clean Third Normal Form, beautiful foreign key constraints, and column names that actually describe what’s inside them. You fire up your connection, query the information schema, and… Silence. Zero foreign keys. No primary key constraints. Table names like xref_p10_not_deleted

Let’s have an honest conversation around the digital water cooler. How many hours of your life have you spent writing regular expressions to try and salvage a piece of JSON that an LLM mutilated? You know the drill. You ask GPT-4 for a nice, clean Python dictionary of customer sentiment scores. It thinks for a while, spits out 500 tokens of absolute gold, and then… forgets the final closing curly brace. Or maybe it adds a trailing comma that chokes Python’s json.loads() like

Here at Factal, we have spent the better part of the last few years whispering sweet nothings into the ears of Large Language Models. We, like you, learned the magic incantations. We learned that if you want GPT-4 to solve a math problem or analyze a messy dataset without hallucinating a conspiracy theory, you have to use the Golden Prompt: "Let's think step by step." It was the WD-40 of prompt engineering. It fixed everything. It turned "System 1" rapid-fire guessers into "S

At Factal, we’ve decided that standard schema mapping is for amateurs. We're currently knee-deep in a challenge we call "The Redundant Relative." In complex ecosystems like Salesforce, objects are often related in ways that would make a soap opera writer blush. Let's consider the case of the hypothetical Timesheet Detail object, the one that inspired our last post. It might have a direct lookup to an Account, but it also  links to a Case, which - surprise! - has its own looku

This image has nothing to do with the post, our founder is just in Cambodia right now 😉 If you’ve ever tried to write a SOQL query and thought, "I’ll just group these results by Case Number, because I am a rational human being," Salesforce has a very special middle finger reserved just for you. In the world of Salesforce, the CaseNumber field is a bit of a peacock. It looks like a string, it acts like a string, but the moment you try to use it in a GROUP BY clause, Salesforc

I spend WAY too much time on Reddit, and recently, I've been noticing an interesting type of post on the ChatGPT subreddit: people are asking GPT 5.2 how many Rs are in "garlic". Way back when (i.e. mid-2024), people used to do a similar trick where they asked GPT 4o how many Rs were in "strawberry", and they got similar results - both these LLMs would regularly come back with the wrong answer (apparently, the Venn diagram of people who like to cook and people who like to mes

Every few decades, the tech industry rediscovers a universal truth: the future is always more expensive than we thought it would be. We’ve hit that point with AI. Right now, we’re still drunk on the miracle of it all - chatbots that reason, copilots that code, agents that talk. It’s intoxicating. It also costs a small fortune to run. The dirty secret is that almost every “AI company” in your feed is currently subsidized by venture capital or by model providers who are themsel