Mark Conway

Quartr API datasets are now available via Snowflake. Access structured earnings data – live audio, real-time and historical transcripts, filings, slide decks, and event summaries. Pull Quartr data directly into your existing Snowflake workflows and augment your datasets. No new pipelines, no separate API integration. Build with data that drives decisions.

75

We've launched the dbt_semantic_view package, seamlessly integrating the Snowflake semantic view with dbt's disciplined workflow. This means you can now: - Centralize Metrics: Define key metric logic (FACTS, DIMENSIONS, METRICS) natively in Snowflake's semantic view for a single source of truth. - Enable CI/CD: Apply dbt's strengths (declarative YAML, testing, and version control) directly to your semantic layer definitions. - Ensure Governance: Automatically manage privileges and validate relationships to ensure every consumer, from BI tools to AI agents, uses trusted data. - Unify your governance framework and ensure consistent, high-performance metrics across your enterprise. Learn more and get started: https://lnkd.in/eZqiwT-g

492

2 Comments

🔍📈 Optimize your search without the manual grind. Tuning search algorithms is complex, but Snowflake Cortex Search makes it effortless. Instead of spending hours fine-tuning parameters, leverage LLMs as judges to generate high-quality evaluation data sets—automating relevance assessments and improving search quality at the click of a button. 🔹 Goldens: The key to measuring search relevance, now enhanced with LLM-generated labels 🔹 Autotuning: Improve search performance without manual intervention 🔹 Streamlit Eval Tool: Evaluate and optimize search with AI-powered insights Start optimizing with Cortex Search today: https://lnkd.in/gEgUPfaA

55

One of the most transformative innovations powering today’s GenAI revolution is the advancement of foundation models (FMs) — large models pre-trained on Internet-scale data that demonstrate remarkable generalization across a wide range of tasks. FMs have redefined performance benchmarks in many fields including, natural language processing, computer vision, and computational biology, becoming the cornerstone of modern GenAI. Yet, progress on structured enterprise data such as tabular time series has lagged behind, where task-specific models optimized on manually extracted signals still dominate. Inspired by the success of FMs in other domains, we set out to explore whether FMs can also be built for finance and banking. After over three years of R&D, I’m excited to share that we’ve deployed our first structured-data FM into production at TD: TD AI Prism https://lnkd.in/gEAb7Scm. Trained on billions of customer events, Prism generalizes to many downstream predictive tasks, delivering higher accuracy and significantly faster time to market. Through this work, we're beginning to observe early signs of scaling laws, similar to those seen in LLMs, where more data and larger models lead to stronger generalization. As part of this effort, we also released a public FM trained on Internet tabular data, capable of making predictions on new datasets without any fine-tuning. Check it out here: https://lnkd.in/gmbWdAf7 As enterprises evolve with AI, we believe that foundation models for structured data will be a major force in redefining how predictive AI is deployed at scale, driving the next wave of innovation in enterprise applications. Layer 6 AI has multiple exciting initiatives in this space, so stay tuned for more updates!

359

15 Comments

Just getting started with LLMs? Don’t let the jargon slow you down. This 5-part LLM Glossary is your shortcut to mastering the fundamentals👇 1️⃣ Model Types • Foundation Model: Pretrained on large-scale, diverse datasets to learn general capabilities. • Instruction-Tuned Model: Further trained to follow user instructions precisely. • Multi-modal Model: Handles multiple input/output formats like text, images, and audio. • Reasoning Model: Optimized for logic, problem-solving, and step-by-step thinking. • Small Language Model (SLM): Lightweight models for fast, efficient, on-device tasks. 2️⃣ Training & Fine-Tuning • Pretraining: Initial learning phase using massive, diverse data sources. • RLHF (Reinforcement Learning with Human Feedback): Aligns model behavior with human preferences. • DPO (Direct Preference Optimization): Uses ranked preferences instead of rewards for training. • Synthetic Data: AI-generated data used to supplement real datasets. • Fine-Tuning: Targeted retraining on specific domains or tasks. • LoRA / QLoRA: Techniques for efficient low-resource fine-tuning. • Guardrails: Rules or filters applied to enforce safety, ethics, and compliance. 3️⃣ Prompt Engineering • System/User Prompts: Define model behavior and user context in a session. • Chain-of-Thought (CoT): Prompts that guide the model to reason step-by-step. • Few-Shot / Zero-Shot Learning: Demonstrating tasks with few or no examples. • Prompt Tuning: Optimizing prompts via training for specific outcomes. • Context Window: The total number of tokens the model can "remember" per interaction. 4️⃣ Inference & Generation • Temperature: Adjusts output randomness; lower values are more deterministic. • Max Tokens: Limits the number of tokens in a generated response. • Seed: Controls reproducibility of outputs. • Latency: Time taken by the model to respond. • Hallucination: When the model generates factually incorrect but plausible-sounding information. 5️⃣ Retrieval-Augmented Generation (RAG) • Retrieval: Fetching relevant data from external sources before generation. • Semantic Search: Search based on meaning, not just keywords. • Embeddings: Vector representations of text used for similarity matching. • Chunks: Dividing documents into smaller segments for better retrieval. • Vector Databases (VectorDBs): Stores embeddings for fast and accurate search. • Reranking: Rescoring retrieved documents to prioritize relevance. • Indexing: Structuring data for efficient retrieval during generation. 🌟 Follow the AIKaDoctor (Free AI & Data Science Resources) channel on WhatsApp: https://lnkd.in/dCTCEKKc 📌Follow Habib Shaikh For more such content.

232

36 Comments

We open-sourced our framework for validating #ExplainableAI methods! Samuel Sharpe, Sr. Lead Machine Learning Engineer, shares how his team developed guardrails for ablation studies on tabular data. Check out the framework and test your own XAI methods. https://bit.ly/4esirrV

43

2 Comments

RAG isn’t just a buzzword—it’s about using semantic search and AI to find and use information smarter and faster. Rini Vasan, an AI Product Marketing Manager at Redis, set out to build a Retrieval Augmented Generation (RAG) pipeline using the Redis Vector Library and also created a working AI assistant. Learn about her experience with RedisVL and start yours today: https://lnkd.in/gj2T4xsU 

74

🚨 Avoiding Common Pitfalls in Productionizing LLMs: An AI Architect's Experience 🚨 Building real-world generative AI? Here's what most teams miss—and how to avoid the classic mistakes: Why LLM Projects Fail: - Only 22% data accuracy in production (financial services) - Code often gets reverted after launch - Large enterprise tools underdeliver Top Technical Challenges: - Model serving: optimize for speed, reliability, and resource use - Hybrid architectures add complexity - GPU/TPU acceleration and batching boost performance - Version control & AB testing keep deployments sharp Cost Optimization: - Track token usage and set budget alerts - Use hybrid cloud and monitor resources - Smart model routing can save big (up to 60-70%) - Caching & model distillation reduce bills Quality & Reliability: - Use RAG and vector search for accuracy - Monitor hallucinations, consistency, satisfaction - Continuous retraining to tackle non-determinism Security & Compliance: - Guard against prompt injection, PII leaks, bias, toxicity - Meet all key regulations (GDPR, HIPAA, SOC 2, etc.) - Enforce validation, filtering, and security Scale with Confidence: - Model parallelism & Kubernetes orchestration - Edge deploy for ultra-low latency - Monitor TTFT, throughput, error rates, GPU/memory, cache Deployment Roadmap: - Tight KPIs and basic monitoring to start - Iterate with CICD and strong guardrails - Use feedback, cost tracking, phased rollouts to optimize Core Principles: ✨ Start small, scale smart   💡 Design for cost-efficiency   🛡️ Quality is non-negotiable   🔒 Security everywhere   👀 Let data guide decisions  Ready to build robust, scalable, cost-effective AI systems? Let's connect! #AIPitfalls #LLMDeployment #AIArchitect #ProductionAI

89

4 Comments

🧠 The Death of the Predictive Model Why LLMs Are Replacing Scores With Reason For years, predictive models sat at the heart of enterprise decision systems. Give them structured data and a well-labeled outcome, and they’d offer what we wanted: a score. Fraud probability = 0.92 Churn risk = 0.81 Approval likelihood = 0.67 And we built pipelines, dashboards, and workflows around those numbers. But models never knew why they were confident. And they never asked if they should be. ⸻ I. Predictive Models Were Born of Constraint They made sense when: • Data was clean and labeled. • Decision latency mattered more than explanation. • Feature engineering was a craft. But they were static, narrow, and deeply reliant on historical assumptions. They didn’t adapt easily. They didn’t narrate. And they didn’t know when to stop. ⸻ II. Why They’re Cracking Now • Real-world patterns shift faster than retraining cycles. • Feature drift quietly erodes accuracy. • Business teams demand explanations—not just scores. • And regulators are increasingly asking: “How do you know this model should have acted?” Prediction pipelines weren’t designed for self-awareness. They assumed that if a score could be computed, it should be trusted. ⸻ III. LLMs Are Rewriting the Decision Layer Large Language Models (LLMs) aren’t just chatbots—they’re reasoning layers. LLMs don’t simply predict; they: • Interpret behavior • Simulate alternatives • Surface confidence • Explain why a decision does—or doesn’t—make sense And most importantly, they can say: “I don’t have enough signal to make this decision.” That’s something predictive models never learned to do. ⸻ IV. Stripe’s Example: Adding Reason to Prediction Stripe’s fraud detection system is a case in point. In 2024, they introduced a transformer-based Payments Foundation Model, trained on billions of transaction sequences. It was designed not to replace scoring models, but to augment them with pattern recognition, sequence reasoning, and contextual understanding. The result? Detection of certain fraud attacks (like card testing) improved from 59% to 97% overnight—not by replacing the predictive core, but by surrounding it with better reasoning and input interpretation. This shift reflects a larger trend: Prediction is being unbundled. Reasoning is taking the lead. ⸻ V. Even Tabular Isn’t Safe This isn’t just about unstructured data or documents. LLMs are making structured, tabular data more intelligible too. Instead of asking, “What’s the fraud score?” We can now ask, “Is there enough context to decide, and if so, why?” ⸻ The new core is a system that can: ✅ Reason through inputs ✅ Reflect on confidence ✅ Decline when unsure ✅ Explain when it proceeds ⸻ Predictive models gave us numbers. LLMs give us judgment. #tPower #mlai #GenAI4FS #inc81starch

71

27 Comments

Building agentic workflows? There are actually only a few key concepts to be aware of. StateGraph orchestration: A StateGraph is the agent’s blueprint. It consists of nodes that handle tasks with edges (directions) to other tasks, defining the order. Together, they provide the agent with a flow to follow. Conditional routing: At decision points, the agent inspects results and follows “if-else” statements that skip failures, handle edge cases, etc. Memory and threads: Thread memory holds recent conversation turns. Essentially, it stores the short-term memory that you are John Doe and were interested in the weather. Long-term memory stores user facts, letting the agent stay context-aware, reduce repetition, and personalize replies, such as that you have been a customer since 2018. Human-in-the-loop: Strategic hand-offs let people review low-confidence outputs, catch critical paths or errors, and feed corrections back to the model, keeping quality and alignment tight. If this still feels too abstract, I’ve included a more detailed explanation in the article below. https://lnkd.in/eAr-6YAQ

13

1 Comment

Brilliant post by Jessica Talisman on the difference between semantic layers and ontologies. The frame couldn't be clearer: - Semantic layers solved for metric governance in a world where users interacted with dashboards - Ontologies provide a structured representation of knowledge where AI systems can access context and make inferences Full article here: https://lnkd.in/eBDyMUfa

14

Evaluating LLMs goes beyond simple benchmarks. Building reliable and safe AI requires a complete view of model quality that combines quantitative metrics, automated LLM judges, and human evaluation for context and nuance. Learn how leading teams assess performance, safety, and reliability to build trustworthy AI systems: https://lnkd.in/gjJ-ySyV

106

3 Comments

🔍 Why the Future of AI Research Isn’t Just About Retrieval—It’s About Deep Reasoning Large Language Models (LLMs) are no longer satisfied with simply fetching information. Enter Deep Research (DR) Agents—a new class of autonomous AI designed to tackle complex, multi-step research workflows that blend: ✅ Dynamic Reasoning (thinking step by step) ✅ Adaptive Planning (adjusting strategies on the fly) ✅ Tool Use (coding, analyzing data, browsing the web) ✅ Structured Report Generation Unlike traditional retrieval-augmented generation (RAG), DR agents don’t just retrieve and repeat—they synthesize, validate, and plan. What sets them apart? • Hybrid Search: Combining API calls with live browser exploration to handle both structured and messy, real-time content. • Multi-Agent Collaboration: Specialized agents work together, mimicking research teams. • Non-Parametric Learning: They evolve capabilities without retraining, learning by storing and reusing prior workflows. • Dynamic Workflows: From clarifying your intent to planning and executing tasks, all in real time. This paradigm shift has big implications: 🏥 Healthcare: Faster synthesis of the latest studies. 💼 Finance: Real-time risk assessments with live market data. 🧪 Scientific Discovery: Automated literature reviews and hypothesis generation. The paper also highlights key challenges—like aligning evaluation benchmarks with real-world needs and optimizing multi-agent coordination. How soon do you see DR agents becoming your co-pilot in complex decision-making? #AI #DeepResearchAgents #LLMs #FutureOfWork #AIResearch

41

12 Comments

🚀 Context Engineering: The Science Behind Smarter LLMs 📄 Just finished reading “A Survey of Context Engineering for Large Language Models” — and it’s a game-changer for anyone working with or building LLM systems. 💡 Key Takeaways: 🔹 Context Engineering is more than prompt design. It’s the formal optimization of the information payload that LLMs consume — including retrieval, formatting, memory, and orchestration. 🔹 A Unified Framework: The paper breaks down Context Engineering into 3 foundational components: 1. Context Retrieval & Generation – e.g., Prompting, RAG, Dynamic Assembly 2. Context Processing – e.g., Long-sequence modeling (Mamba, FlashAttention), Self-Refinement (Reflexion, Self-RAG) 3. Context Management – e.g., Compression (H2O, QwenLong), Memory Architectures (MemGPT, MemoryBank) These are then integrated into powerful systems: • RAG 2.0+: Agentic RAG, GraphRAG, Modular RAG • Memory-Augmented Agents: Like MemOS and RecMind • Tool-Integrated Reasoning: Think Toolformer, AutoGen • Multi-Agent Systems: CAMEL, SagaLLM, CrewAI 🔍 Why it matters: • 18x gains in information navigation • Up to 94% task success rates using context optimization • Models learn, reflect, and self-improve without retraining • Enables real-time tool use and long-term memory across tasks ⚠️ Biggest Challenge Identified: Despite advances in understanding complex context, LLMs still struggle to generate equally complex, structured long-form outputs. Bridging this gap is a priority for future research. 🌐 What’s Next? • Scalable architectures with context compression • Unified benchmarks for evaluating tool use and memory • Human-AI collaboration powered by multimodal, structured, and self-optimizing systems If you’re building context-aware LLMs, designing retrieval strategies, or exploring multi-agent orchestration — this paper is a must-read. 🔗 arXiv Link: https://lnkd.in/efyyXGQz 📚 GitHub Resource: https://lnkd.in/eiqdA4Dw #LLM #AI #ContextEngineering #RAG #Agents #Toolformer #AIResearch #NLP #MemorySystems #MultiAgentSystems #PromptEngineering #SelfRefinement #GenerativeAI

166

4 Comments

Ever feel like today’s AI models are getting smarter but somehow less useful? You’re not imagining it. LLM benchmarks have become the obsession. Model makers are locked in a cycle of chasing leaderboard points on MMLU, GSM8K, and other tests no one actually cares about outside of research labs. The result? Models that ace exams but fall apart in real-world use. They misinterpret vague instructions. They forget what you said three messages ago. They can’t think beyond a single step. But sure, the score went up by 0.3 percent. Benchmarks were supposed to be a guide. Instead, they’ve become a crutch. A way to show “progress” without improving anything that matters. The questions we should be asking are simple: Can the model follow messy, human instructions? Can it reason through problems, not just regurgitate answers? Can it actually help people get real work done? If you’ve worked with these models, you’ve felt the gap. You’ve seen the hype, and then you’ve seen it fall short when it counts. Are we building for people, or are we building for points? I’d love to hear what others are seeing in the wild. #AI #LLMs #Benchmarks #MachineLearning #RealWorldAI

27

3 Comments

We just released GutenOCR: a vision language model built specifically for OCR, with open code, open weights and open data. Try it here: https://ocr.roots.ai/ We have repeatedly seen this in production document workflows: models can read text, but become unreliable when you ask them to ground that text to a specific location on the page. GutenOCR is our effort to solve this more directly, trained using an open data recipe so the approach is transparent and reproducible. - Paper: https://lnkd.in/epV8MmcH - Code: https://lnkd.in/e4gWuvke - 3B: https://lnkd.in/eKxnhiwP - 7B: https://lnkd.in/euz568DS P.S. Please join our researchers for a live talk + Q&A on how we built GutenOCR. Register here: https://lnkd.in/ecqQUqJP

589

48 Comments