Natural Language Processing news and analysis on Web Pulse: 5446 curated articles tracking the latest Natural Language Processing developments, tools, and research, updated continuously from vetted sources.
Granola, an AI notepad, now enables users to draft product requirement documents (PRDs) and proposals directly from discovery calls by integrating with AI tools like Claude and Cursor via the Model Context Protocol (MCP)…
Granola, an AI notepad company, announced on September 4 that its Granola Chat feature enables users to search across all past meetings using semantic understanding, retrieving exact customer quotes with source-linked ci…
Digital Applied published engineering guidance on September 7, 2026, proposing a review workflow for AI translation QA that separates commercial invariants—such as prices, dates, conditions, and links—from linguistic flu…
A developer detailed common pitfalls in scaling Retrieval-Augmented Generation (RAG) systems, including retrieval latency, data management, and data quality, and offered solutions such as using FAISS for approximate near…
A study of two independent extractors parsing Missouri statutes found a 0.43 false-negative rate on numeric-threshold presence, and a survival analysis using 1,000 Monte Carlo trials with a Wilson 95% lower bound of 0.95…
Conversion, a marketing intelligence company, announced it built a self-improving text-to-query agent that converts natural-language audience descriptions into editable filters, using a tool-using agent, an intermediate …
A developer built a pipeline that translates podcast episodes into other languages for about a dollar per episode, using the podcaster's own published transcript rather than speech-to-text to avoid compounding errors. Th…
MAI-Transcribe-2, a new speech recognition model from MAI, claims to be the fastest, most accurate, and cheapest in the world, ranking first on the FLEURS benchmark across 60 languages with an average Word-Error-Rate of …
A developer has published a technical guide for building a multilingual voice AI sales agent for ecommerce, combining Vapi's voice gateway, Shopify's GraphQL API, OpenAI's GPT-4 for conversation, and Google Cloud Text-to…
Google announced Gmail Live, Docs Live, and a revamped voice mode for Google Keep at its I/O event on May 19, all powered by Gemini AI, enabling users to search emails, draft documents, and organize notes via natural con…
Google is rolling out Gmail Live, Docs Live, and Keep Live this week, bringing voice-based AI interactions to its productivity apps. Gmail Live allows natural spoken conversations with your inbox, Docs Live serves as a h…
EvidSeek, a new browser extension launched on Hacker News, analyzes online discussions and reviews from Reddit, X, Product Hunt, Hacker News, and Trustpilot to generate structured summaries, pain points, unmet needs, opp…
Voice AI agents are emerging as a viable alternative to screen-based enterprise software, offering hands-free interaction that reduces learning curves and improves efficiency for field workers, support, and healthcare. T…
Researchers released VibeVoice-ASR-Streaming, an LLM-based end-to-end streaming speaker-attributed automatic speech recognition system that processes fixed-size audio chunks with lookahead to output 'who said what' in re…
DronaHQ published a step-by-step guide on setting up an outbound AI voice agent for customer service, using a post-service follow-up call as an example. The guide details connecting a phone number via SIP trunk, creating…
Browser-native OCR eliminates the need to upload documents to third-party servers, according to a technical post about @stabrise/scaledp@0.1.1, which runs PDF rendering, text detection, OCR, and entity recognition entire…
A new multilingual summary meta-evaluation dataset (BASSE) containing human judgments on 2,040 abstractive summaries reveals that proprietary judge LLMs correlate most strongly with human ratings, followed by criteria-sp…
Researchers introduced EmoStance, a method that uses emoji distributions as weak supervision to control the affective orientation of empathetic response generation, achieving a 62.2% decisive win rate in blind pairwise e…
Researchers propose Prior-Guided Tuning (PGT) with Contrastive Prior Steering (CPS), a training method that uses task-level natural-language priors as auxiliary learning signals for low-resource LLM training. On AmbiMath…
Researchers introduced TALKFA, a unified benchmark for Farsi dialogue generation and understanding, comprising three datasets: WIKI-FADIAL (4.2K Wikipedia-grounded dialogues), DAILYDIALOG-FA (6.6K dialogues with dialogue…