 |
We Love Ugly Data! The Deep Analysis PodcastAuthor: Alan Pelz-Sharpe
This irregular podcast series examines what is happening in the unstructured data automation market. Three topics - Thirty Minutes, that's the format!Topics range from the state of Blockchain, IDP, ECM, and the impact of AI on unstructured data. Deep Analysis provides advisory services, industry research, and M&A guidance. info@deep-analysis.netDeep-Analysis.Net Language: en-us Genres: Business, Management, Technology Contact email: Get it Feed URL: Get it iTunes ID: Get it |
Listen Now...
S05E10 Headless Salesforce & New Research
Thursday, 8 October, 2026
The forty-fourth episode of the podcast you know and love, "We Love Ugly Data!", is out, available in audio everywhere you get your podcasts and as a video on YouTube (embedded below). Alan's back with Matt this month to talk a few more of the latest drop of vendor profiles, making sense of the big announcements from Salesforce's Dreamforce conference and his choices in our new-ish segment; "Hugs or Bugs".In this month’s episode:Topic 1: New Research (pt3)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 (which has since been updated). This time, Matt and Alan discuss the last two in that collection.First up is TCG, a process automation and orchestration platform that integrates documents, AI, human‑in‑the‑loop, and system‑to‑system interactions. They're specialists in very document-heavy use cases, like you'll find in the insurance industry, and can handle the complexity and document size that comes with processes like claims handling (including document dossier handling, where you're looking for the capability to handle collections of related documents as part of a single instance of a process). Alan points out that this marks TCG out from many of the other IDP vendors that are in the market; Matt mentions that its ability to use a range of different AI models and tools within a single document is helpful in managing their overall transactional costs of processing (remembering that not all models are equal in terms of their cost of operation). Finally, UAI Labs is an automation and AI services company that works with private equity firms to help them better understand the operations of the portfolio companies they invest in, then helps those companies improve their ways of working. The company offers a mix of consulting and technology as a contracted service, designed to help PE shops quickly understand how their newly acquired assets operate and how to improve those operations to increase value. As our latest M&A update notes, it's an interesting market right now for PE firms and specialist tools to help them overhaul and move on firms, which, you would imagine, would be of significant interest in that light. UAI Labs have also licensed the report and made it available free of charge via their website, so you can read all the details without being a Deep Analysis subscriber. Thanks for that.Topic 2: HeadlessforceLast month was Salesforce's Dreamforce conference; Alan was on the ground in San Francisco, Matt was in his home office in the UK, and the pair published a blog post wrapping up their early analysis. Here, they discuss AI Force, which Alan wrote about during his visit to the Salesforce developer conference TDX back in April, and why any analysis that says "they put the CRM into Claude" with Claudeforce is very dumb indeed. They look at how the headless thing needs a head - and Claude is an example of such a head - to operate and discuss some of the implications of that from a cost perspective. Matt also calls out a couple of good independent, Salesforce-specialists for their work (Salesforce Ben and Apex Hours) explaining some of the related concepts. There's also a discussion of how Agentforce has shifted to a "select your hero, adjust if you must" approach to get customers live as quickly as possible (quite a change from its launch 2 years ago).Topic 3: Hugs or Bugs?It's time for another round of our new podcast game, following on from our previous ones on banned words, analyst value, and briefing dos and don'ts, which Matt calls "Hugs or Bugs?". Here, Alan shares a trend he likes (Hugs!) and one he dislikes (Bugs!). The hug goes to the DM Awards, which Alan has judged for years (and is currently judging the 2026 version), because it offers great insight into real-world customer projects. The bug stems from confusing AI with LLMs; Alan is annoyed that people assume all AI is tied to the relatively recent arrival of LLMs, rather than the decades of AI technology that came before (for example, all those enterprise-grade ML projects). https://www.youtube.com/watch?v=GjWuN0T1GHk Related Links for Series 5 Episode 10Alan's newly updated summary post of all the latest batch of vendor profilesDownload the UAI Labs report free of charge, courtesy of the company. The big joint Salesforce Dreamforce blog post from Matt and Alan.Apex Hours detailed discussion of how Claudeforce works in practice.Salesforce Ben's story about how Salesforce is changing how they're charging for Agent access.The DM Awards.Support the show
