Adamma Morrison

Latest addition to Trevorwithdata? An amazing AI assistant with a very corny name: Ask Trevor 😆 If you've been following this #journeyanalytics build, you know there are a LOT of things that need to be configured before anyone can start analyzing data - warehouses, datasets, identity stitching, metrics, dimensions, segments. And every step has enough nuance that getting it wrong means your data looks weird downstream 😬 So I built an AI assistant on top of the whole thing. It uses Anthropic's Claude Agent SDK (same framework that powers Claude Code) with 13 custom MCP tools that map to every API in Trevorwithdata. The idea is simple - anything a user can do through the UI, the assistant can do too. Here's what it actually does in practice: you describe your data and your KPIs in plain language, and the assistant handles connecting your warehouse, creating datasets, setting up identity stitching, defining metrics and dimensions - the whole pipeline. It's not just automating clicks. It's making real decisions, like recognizing that "revenue per visitor" needs two standard metrics created before it can build the calculated one, or defaulting to person-scoped segments for cross-device journeys. The secret sauce is honestly a 1,800-line system prompt where I crammed everything I know about journey analytics 😅. Decision rules, workflow patterns, scoping logic - all the domain knowledge that would normally take weeks to learn. The vision goes way beyond just setup (analysis, reporting, exploration are next), but I wanted to show where it was so far. Check out my latest post to see how I built it (and some of the system prompts I created) - link in comments 🔗

35

4 Comments

LLMs and Analytics Everyone's excited about LLMS for writing content. But when you try to apply them to analytics, the challenge is real. LLMs are probabilistic by design. This works great for creative tasks where variation is ok or even wanted. But analytics requires deterministic results. When someone asks "how many units did we sell last quarter?", there's only one correct answer. However, two users can ask the same analytical question and receive different answers. Not because the LLM is broken, but because business questions are inherently ambiguous and LLMs will resolve that ambiguity differently each time. Example from a non-technical user: "What product sold more last month?" There are at least a dozen valid interpretations: "Product" = SKU? Product line? Brand? Category? "Sold more" = Revenue? Units? Profit? Growth rate? "Last month" = Calendar month? Last 30 days? Fiscal period? An analyst knows to ask clarifying questions. A non-technical user doesn't realise they're being ambiguous. And an LLM will happily pick an interpretation, potentially a different one each time. This inconsistency erodes trust. The DATA/AI industry talk about "text-to-SQL" as if it's a solved problem just like that. It's not. Here's what's actually required: Semantic Standardisation: Mapping business language to technical reality. When users say "product," what do they actually mean in our context? Curated Question Libraries: Pre-built patterns for common queries that act as interpretive anchors and ensure consistent interpretation. Layered Context: The LLM needs organisational context (how your business operates), domain context (rules for specific verticals), and data context (what tables contain and how they relate). Progressive Complexity: Layering is paramount with simple counts and sums. Then, introducing ratios and trends. Each layer builds trust and teaches better patterns. This means that we have: LLMs for writing: Embrace variation, encourage creativity. LLMs for analytics: Constrain interpretation, standardise assumptions, validate relentlessly. The technology is advancing rapidly, but the core issue remains: Business language is ambiguous, and analytics demands precision. We need to invest in the unglamorous work of mapping business vocabulary to data reality, building interpretive frameworks, and designing systems that build trust through reliability. #ai #analytics #metadata

77

6 Comments

Why MCP Matters for the Future of AI As LLMs move from demos to real-world systems, the biggest challenge isn’t model capability—it’s context, action, and integration at scale. In this blog, I break down Model Context Protocol (MCP): • Why prompt-based integrations don’t scale • How MCP solves the M×N integration problem • Why MCP is emerging as the operating system for AI agents If you’re building agentic systems, RAG pipelines, or enterprise AI, MCP is a concept worth understanding early. 🔗 Read the blog here: https://lnkd.in/gJpcAHdt #AI #LLM #MCP #AgenticAI #Architecture #FutureOfAI

4

Here's a few things I found after playing with commercial and transactional prompts on Perplexity Comet for ~3 hours: > Agents prefer websites with streamlined access to data, and that preference seems to be hardcoded to some extent. In doing 'research + add to my cart' agent shopping tasks, Comet consistently started out by navigating to one retailer's website, even when I stated I preferred it's direct competitor! > One reason for that might be the quality and consistency of the website's data structure. I am thinking back to some of the monster ecommerce retailer sites I have seen. The ones built on a homegrown CMS in 1996, with another CMS bolted on in 2002. New product categories and experiences are added to the site in a way similar to the Weasley's house in Harry Potter. If agentic commerce takes off and requires consistency in data structure, those sites are going to be cooked. > There are at least two clear agentic user flows that brands need to consider. One is a research flow where the agent will go off, do research, and come back with a comprehensive recommendation that mentions many brands and cites many webpages. This is just what SEO will become. > The other is a true virtual assistant flow where, if you know what you want, the agent will find it for you and present you with a filled cart. For these user flows, you may have a customer who's at the finish line ready to convert and gets snatched away at the last second because the virtual assistant found a better price, better stock, or (worst of all) just had an easier time navigating a competitor's site. > Shopping feeds are going to be critically important. If you're an ecommerce brand and your shopping feeds are jacked up, you are not going to have a good time acquiring customers on search. It's easy to say "who cares about Perplexity Comet, they have no market share" but I would caution against that mindset. Google is laying the foundation to enable the same experience in their ecosystem. Comet should be the sandbox you test in to prepare for Google's agentic browsing capabilities.

17

5 Comments

AI is now the fastest‑spreading general‑purpose technology in history, reaching over a billion users in under three years. The Microsoft AI Diffusion Report shows that this momentum is reshaping economies, with adoption highest in countries that have invested early in digital infrastructure, skills, and supportive policy environments - countries like Singapore, the UAE, Norway, and Ireland are standing out in this regard. These “diffusion leaders” demonstrate how quickly AI can scale when the fundamentals are in place The report also highlights widening global gaps. AI use in the Global North is nearly double that of the Global South, and language has emerged as a new structural barrier. Countries where low‑resource languages dominate see significantly lower adoption even when connectivity and income levels are similar. Meanwhile, AI infrastructure remains highly concentrated, with the U.S. and China hosting most global compute capacity, though several countries are rapidly closing the frontier‑model gap The key takeaway is clear - the next phase of AI progress will be defined not by who builds the most advanced systems, but by who ensures broad, equitable access. Diffusion, not invention alone, will determine whether AI becomes a driver of inclusive economic growth. Expanding infrastructure, skills, and language coverage will be essential to making AI’s benefits truly global - See the full report at https://lnkd.in/euAwn5v8, and below an interview with our Vice Chair and President Brad Smith #diffusion #AIleadership #microsoft

22

There are very few things that excite me more than applying my experience in new ways. The Master of Analytics program in NC State University allows me to do just that, applying lessons I've learned in intelligence research to data science. While there are endless topics in #machine_learning I am eager to write about, one topic that felt especially close to heart was how language is shaping our decision-making. In "Why Lingo Matters in Analytics", I chose to tackle one specific aspect of this: how our framing of the data must be as accurate and formal as our actual data. Our framing must be accurate and formal, because all throughout history people were hearing and reading ideas via *words*, not numbers; and thus we evolved to make decisions based on *words* and not based on numbers. When a data analyst/scientist makes a recommendation to a stakeholder, the actionable part is always in the words, never in the numbers. The numbers are there only to show rigor and make our claims verifiable. To quantify. To answer the "how much", but never the "what" or "why". The true benefit of AI is allowing us to reach conclusions much faster (and make funny animal videos faster too, I guess). But to reach these conclusions safely, and make the right decisions, we need the right framing to judge the numbers by. Have a fun read!

11

AI & Market Research - Can AI take surveys for us? I haven't posted about too many uses of AI with Market Research to date, however I think this case study should get a mention. A recent study by PyMC Labs and Colgate‑Palmolive finds that large language models (LLMs) can closely simulate human responses in consumer product surveys—hitting ~90% of human test–retest reliability across 57 surveys (9.3k human responses). How it worked: Instead of asking a bot for a number, the team asked for a short written answer with a why. That text was then compared to five Likert “anchor” statements using embedding similarity (their Semantic Similarity Rating / SSR method). Result: 90% of human test-retest reliability. They mirrored distributions of responses across demographic segments such as age and income. Researchers emphasise that this doesn’t mean LLMs think like people, but their outputs can align with human behavioural patterns under defined conditions Want to try this? 1) Start with a low-risk, high-volume concept test you already run. 2) Write clear Likert anchors (1–5) as plain sentences. 3) Prompt an LLM for a short rationale per concept and then map that text to anchors using embedding similarity. 4) Create a small human holdout to benchmark; only proceed if your synthetic vs human gap is tight. 5) Be transparent: label synthetic vs human, document prompts, and keep humans in the loop for final calls. Imagine being able to test new tourism or retail products or ad concepts instantly across synthetic “audiences”, before spending on full panels. Huge potential for insight and speed — as long as we keep the human layer where it matters most: judgement, context, and creativity. This isn’t about replacing people; it’s more about "simulate → iterate → validate" so teams can test more ideas, faster, and spend human budget where judgement matters most. Full story below 👇 #AI #MarketResearch #Insights #DigitalTransformation #SyntheticData #Tourism #Innovation https://lnkd.in/eRBWtRcs

6

1 Comment