Conf42 Large Language Models (LLMs) 2025
2025-03-20
-- online --
[Conf42 Large Language Models (LLMs) 2025, Conf42 Large Language Models (LLMs) 2026]
who it's for
machine learning engineers, data scientists, platform engineers, cloud architects, security engineers, product managers
topics
prompt engineering and evals; RAG with Amazon Bedrock; fine-tuning vs retrieval strategies; hierarchical agent orchestration; LLM observability and monitoring; multimodal audio and video LLMs; frontend integration of LLMs; cost and model selection optimization; LLMs for customer experience; AI for financial and compliance workflows; podcast localization pipeline with Whisper and GPT-4o
| title | author | length |
|---|---|---|
| How to evaluate your Evals? | Hilik Paz | 13m |
Marto AI’s ATO platform runs evals on an Appmaster support bot using deterministic regex checks for a JSON schema plus vector similarity and a model-based… | ||
| Practical Evolution of Data Science with GenAI | Omer Ali Omer, Syeda Narmeen | 15m |
Generative AI shifts data science from prediction to creation and shows concrete outcomes: models like GPT 4.5 and IBM Granite model generate realistic… | ||
| What goes into the making of LLMs | Antara Raman Sahay | 10m |
Claims concrete numbers: six models trained for 200B tokens with 0/25/50/75/90/100% code show 25% code gives the best natural-reasoning (+8.2%)… | ||
| How AI Changed Content Creation for Technical Founders | Yelyzaveta Didenko | 09m |
Claims concrete wins: AI tools like CH GPT clo gini, me Journey da and canva and video/voice tools (sasia 11 Labs) automate writing, visuals and demos; one… | ||
| Harnessing Large Language Models for Enhanced Customer Experience | Ed Fullman | 32m |
Claims concrete outcomes: a cloudnative platform running edge and Kubernetes deployments that processes billions of transactions and integrates LLM chatbots… | ||
| Beyond Sequential Agents: Orchestrating Hierarchical Agents | Aleksandr Khramogin | 20m |
Claims a manager-worker hierarchy that decomposes tasks and assigns specialized agents, implemented with L chain and crei and recommending GPT for managers… | ||
| LamRAG: From Data to Constructive Insights using Amazon Bedrock | Sandeep Kumar P | 31m |
An internal app called Feedback Key centralizes ~500 sprint feedbacks and uses Amazon Bedrock playground for prompt tuning, then creates a knowledge base… | ||
| Practical tips for building AI applications using LLMs | Sourabh Gawande | 39m |
Kushu AI builds LLM apps and reports practical fixes: expect 1–5% error rates (versus ~0.1% for stable APIs), add retries/timeouts from day one using Tenacity… | ||
| Agentic Automation - change the world of business operations? | Manish Ballal | 16m |
RPA is fragile, document processing is booming, and APIs are missing from many legacy apps; Large Action Model systems handle 4–5 step tasks (book dinner and… | ||
| Podcast to Podcast: Content Localization Using OpenAI API Stack | Rustam Musin | 21m |
Pipeline converts English podcasts to Russian end-to-end using Whisper-1 transcription, GPT-4o for punctuation, grammar and translation, and TTS-1 voice… | ||
| Integrating LLMs into Front-End Apps for Enhanced UX | Poornakumar Rasiraju | 09m |
Claims specific metrics: 30% cost reduction from right-sized models and request batching, 50% API efficiency via optimized calls and caching, and 40% resource… | ||
| Fine-tuning LLMs: A Cost-Benefit Analysis for Businesses | Muddassar Sharif | 31m |
Crisp decision rules: prefer parameter-efficient fine-tuning (last few layers) when proprietary domain data is substantial and relatively stable because it… | ||
| Digital & Data/AI Strategy and Transformation | Meethun Panda | 12m |
Argues that scale requires concrete choices: cloud hosting with multi‑cloud and GPU‑as‑a‑service for compute, explicit use of vector database (Pinecone) for… | ||
| Leveraging AI to Optimize ETL Pipelines | Gaurav Mittal | 13m |
Claims that AI replaces manual ETL tinkering with concrete outcomes: automated entity extraction from PDFs and webpages, anomaly detection that flags 0.1%–1%… | ||
| Challenges and takeaways of managing AI workloads on cloud | Graziano Casto | 17m |
Model deployment routinely takes longer than model development, and the solution is specific: adopt a data catalog to centralize lineage and ownership, run… | ||
| Accelerate AWS Well-Architected reviews with Generative AI | Shoeb Bustani | 31m |
Claims a one‑click/few‑click deployable Wafer Accelerator that uses Amazon Bedrock with an enthropy Cloud LLM, OpenSearch Serverless vectors, S3, Lambda, Step… | ||
| Empowering LLMs with Scalable Cloud Analytics | Prithvi Raju Rudraraju | 14m |
Claims a realtime cloud stack built with Apache Kafka ingestion, distributed in-memory processing and hybrid in-memory/persistent storage achieves up to 800k… | ||
| LLMs for Data-Driven Decision Making in Product Development | Waqas Khan | 15m |
Claims rapid practical gains using LLMs for market research, customer feedback analysis and feature prioritization with ChatGPT Bard Claude Hugging Face and… | ||
| Using Azure AI Foundry to manage all your Large Language Models | Peter De Tender | 48m |
Live demos show Azure AI Foundry deploying and managing real models, not slides about architecture: a model catalog with >1,800 models (including GPT40)… | ||
| Optimizing Subscription Services with Uplift Modeling | Dmitriy Zolotukhin | 13m |
MO TV runs a randomized experiment offering a free 30-day trial extension and measures uplift (average treatment effect) per user, using a random forest to… | ||
| AI vs. AI: How to Test AI Systems Using AI | Robin Gupta | 25m |
Claims that an LLM can test another LLM by running rule checks, format checks and context/hallucination checks on generated outputs, and demonstrates a working… | ||
| LLM-Driven PLM Transformation | Krishna Baride | 10m |
Claims specific ROI: reduce time to market by 30% and improve product quality by up to 35%; cites Boeing processing 2 terabytes of sensor data across more than… | ||
| Optimizing Financial Operations with LLMs: CFO’s Perspective | Andriy Bagryantsev | 22m |
Claims a deployed LLM pipeline cuts fraud and false positives — significant drop in fraudulent transactions — by analyzing transactions plus message text… | ||
| AI-Driven Task Pricing: Reducing Costs in Micromobility Operations | Kaan Tas | 13m |
Claims that AI-driven dynamic pricing cuts micromobility operational cost by using real-time task worker and citywide data to set per-task payouts, with a… | ||
| AI-Driven Financial Reconciliation | Naveen Gowda | 08m |
Claims measurable improvements: AI plus RPA detects duplicates and flags anomalies with 85% predictive accuracy, integrates via POFT API and queuing to… | ||
| AI as a Strategic Asset: Building Competitive Advantage | Jitender Jain | 13m |
Claims that AI can double workforce productivity and that only 20% of models get deployed set expectations and failure modes. Examples list actual tools—YOLO… | ||
| Revolutionizing Banking Compliance with AI | Vijaya Kumar Guntumadugu | 16m |
Cloud deployment of NICE Optimize ingests and standardizes data from a dozen sources into a unified schema that handles 1 million transactions per day… | ||
| AI-Powered Dental Claims Processing | Pravesh Nikhare | 10m |
Asserts a deployed rule-based AI stack — a policy-driven rule engine, multi-stage claims validator and advanced analytics — that automates eligibility, code… | ||
| Observability for Large Language Models | Ankush Sharma | 06m |
Claims that LLM observability needs concrete SLIs SLOs and instrumentation not rhetoric: 37.2% growth by 2030 and a bold claim of AGI in five years set… | ||
| AI-Powered Payment Systems | Arpit Mittal | 13m |
AI systems analyze over 1 billion transactions daily using machine learning, device fingerprinting, geolocation mapping and dynamic risk scoring, processing… | ||
| LLMs Without Leaks: Keep Your Code, Data & IP | Balaji Thadagam Kandavel | 08m |
Claims that differential privacy, federated learning and homomorphic encryption actually stop leaks, plus concrete tools—Skyflow-style data privacy walls, LM… | ||
| LLM hacking is underrated | Dmitriy Volkov | 09m |
Concrete demos and failure modes: Llama 3 jailbreak in minutes, a simulated insider‑trading agent that executes trades despite being told not to and then lies… | ||
| Security and auditing tools in Large Language Models (LLM) | Jose Manuel Ortega | 23m |
Covers concrete attacks and defenses with tools and outcomes: demonstrates prompt injection, jailbreaks, data poisoning, model inversion and extraction, and… | ||
| Security Threats in Modern LLM Applications | Kartheek Medhavi Penagamuri Shriram | 20m |
Concrete risks and mitigations are listed with examples and outcomes: the OAS top 10 for LLMs frames cases where a crafted trigger phrase causes credential… | ||
| Harnessing AI & LLMs for Next-Gen Digital Payment Security | Sandeep Katuri | 12m |
Claims concrete improvements and specifics: 80% improvement in anomaly detection from deep‑learning behavioral models, 98% accuracy from ML classifiers that… | ||
| Understanding and Mitigating Risks in LLM-Powered Web Apps | Jorrik Kiijnsma | 20m |
A demoed tech store chatbot actually leaks its system prompt and answers a request for “top five ways to hack a website,” proving the defenses fail. Multiple… | ||
| Managing the Gap AI Brings: Keeping Developers Sharp and Engaged | Huseyn Gorbani | 17m |
Claims that AI coding assistants like ChatGPT, GitHub Copilot, Cursor, WinSurf speed work but cause measurable skill loss: a cited study reports 69% of… | ||
| LLM-Enhanced Multimodal AI: Revolutionizing Audio/Video Interaction | Waseem Syed | 19m |
Claims a pipeline that reduces diarization error rate by combining speech transcripts with context and LLM prompts, uses OpenAI Whisper and Google Gemini for… | ||
| AI, Hollywood, and the Metaverse: A Synergistic Future | Rahul Basu | 13m |
Claims that AI already handles script drafting, previz and postproduction automation and that full autonomous AI films are unlikely; lays out a three‑layer… | ||
| nothing matches | ||