The forty-third episode of the podcast you know and love as "We Love Ugly Data!" is out, available in audio everywhere you get your podcasts and in video on YouTube (embedded below). Dan's back with Matt this month to talk about a few more of the latest drop of vendor profiles, the great Instabase/SuperApp pivot, and his choices in our new segment, "Hugs or Bugs".

In this month’s episode:

Topic 1: New Research (pt2)

Last month, we started discussing the latest set of vendor profiles made available to subscribers in July; Alan published a blog post with a brief summary of each at the time. This time, Matt and Dan discuss the next three in that collection.

First up is Flexor, which has pivoted from data extraction to providing data with veracity for agentic use cases: the translation layer between documents and agents that fills in the parts LLMs struggle with. Next up is Haystac and its Ominsuite, which is a set of document processing pipeline tools that are largely OEM-ed to other software companies. They use small, industry-specific VLMs (Visual Language Models) that process image versions of documents and can be deployed on an air-gapped, discreet on-premises basis.

Finally, we have SimplAI, an agentic automation AI platform designed to help enterprises and system integrators operationalize agentic AI. This enables the development and management of conversational AI agents, voice and vision functions, and IDP and full multi-step agentic process automation. Dan mentions that this fits into the discussion in his recent blog "Have we reached “peak” AI model wars?". SimplAI has kindly made this vendor profile available from their website for free and has also published a comprehensive blog post about their report here. Thanks to them for that.

Topic 2: Super Pivot?

In August, the heavily backed IDP vendor Instabase announced a significant pivot in its business, deciding it no longer wanted to be an IDP vendor and instead wanted to play in the nascent world of "multiplayer AI" and be called a Super App. Dan has followed the company for a while and wrote about this pivot here, and here reminds us that Instabase never really wanted to be thought of as an IDP vendor in the first place. He also walks us through the various rounds of investment the company received (~$300m; a record in this market) and how the most recent round brought in new investors who may have influenced this change of heart. Dan also reminds us that he's been warning - ever since our predictions in late 2024 - that for IDP vendors to grow to the size investors might wish, they need to grow beyond what IDP currently was ("crossing the border" in his terms). Matt and Dan go on to discuss what this super pivot might mean in terms of the newly named entity's future business and competitors.

Topic 3: Hugs or Bugs?

It's time for a new podcast game - following on from our previous ones on banned words, analyst value and briefing dos and don'ts - which Matt is calling "Hugs or Bugs?". Here, Dan offers a trend which he likes (Hugs!) and one that he dislikes (Bugs!). The hug goes to open-source and open-weight LLMs for how they're enabling innovation for software companies, large and, critically, small. The bug goes to companies that talk about their revenues in ARR terms (Annual Recurring Revenue, so multiplying a recent period to create a yearly figure) rather than TTM (Trailing Twelve Months, so real revenue for the preceding 12 months). Turns out both Matt and Dan have strong opinions on this one. STRONG opinions.

Support the show

The forty-second episode of the podcast that you know and love as "We Love Ugly Data!" is out; available in audio form everywhere you get your podcasts from and additionally in video form via YouTube (which we've embedded below). It's Matt and Alan again this month to talk about the first few of the latest drop of vendor profiles, technology market inertia, and Intelligent Guerrilla Warfare for information managers.

In this month’s episode:

Topic 1: New Research (pt1)

First up, we recently published a big drop of new vendor profiles - 8, in fact - for subscribers. For both subscribers and non-subscribers, Alan published a blog post with a brief summary of each. Here, Matt and Alan discuss the first three in the collection; first up is Doc Labs, the first of 2 AEC specialists in this set. They tackle a real problem: document sprawl in the pre-construction phase, where multiple parties manage their share of the documentation in their own repositories, but there's still a need for overall oversight of all project documents. Next up, also in the pre-construction phase of AEC projects, is Togal.AI, take-off software for estimation that extracts relevant details from plan documents.

Our returning champion amongst this set is Vertesia - which we last covered a couple of years ago - and since then has developed into an ECM-lite platform for AI-enabled marketing content. From its starting point in prompt engineering, Vertesia has built a platform that handles that, as well as model management, agentic development, and their historic skills in content management. If you'd like to read this profile in full, Vertesia has kindly made it available for free download from their website.

Topic 2: Technology inertia?

Matt, Dan and Alan have been planning exciting things for the rest of the year (look surprised when they happen), but as a result of that, they've been discussing what feels like an inertia in technology markets, created by the inability of customers to swallow the big agentic automation products that have been pushed in front of them. Matt's written a couple of blog posts ("Looking fast, but standing still" and "Real world problems remain real, until they’re not") and Alan's been thinking along similar lines from his last visit to Silicon Valley. Here, they discuss why we shouldn't assume automation is always coming to replace inefficient processes; sometimes it's replacing ones held together by real human ingenuity and quality of delivery.

Topic 3: Intelligent Guerrilla Warfare

Alan recently had a speaking slot at a conference in Philadelphia (that he mistook for a nuclear industry bash for some reason). He took the opportunity to provide some hints and tips for information management professionals on how to get themselves involved in AI projects ("infiltrate, don't confront", apparently). Matt also tries a steam train analogy for some reason, which doesn't really land.

https://www.youtube.com/watch?v=VnLiyZ5WUcY

Related Links for Series 5 Episode 8
Download the full Vertesia vendor profile for free from their website.
Here's Alan's summary blog post of this entire set of new vendor profiles.
Matt's blog posts: "Look fast but standing still" and "Real world problems..."
Alan's Silicon Valley thoughts in "Determinism in Silicon Valley".

Support the show