Core Insights - The article discusses the evolution of Agentic AI from pipeline-based systems to model-native paradigms, emphasizing the internalization of reasoning, memory, and action capabilities within the models themselves [1][44]. - It highlights the role of reinforcement learning (RL) as a driving force in transforming static models into adaptive, goal-oriented entities capable of learning from interactions with their environment [1][44]. Background - The rapid advancement of generative AI has primarily focused on reactive outputs, lacking long-term reasoning and environmental interaction. The shift towards Agentic AI emphasizes three core capabilities: planning, tool usage, and memory [3]. - Early systems relied on pipeline paradigms where these capabilities were externally orchestrated, leading to passive models that struggled in unexpected scenarios. The new model-native paradigm integrates these capabilities directly into the model parameters, allowing for proactive decision-making [3][6]. Reinforcement Learning for LLMs - The scarcity of programmatic data and vulnerability to out-of-distribution scenarios necessitate the use of result-driven RL to internalize planning and other capabilities, moving away from prompt-induced behaviors [6][7]. - RL offers advantages over supervised fine-tuning (SFT) by enabling dynamic exploration and relative value learning, transforming models from passive imitators to active explorers [8][9]. Unified Paradigm and Algorithm Evolution - Early RLHF methods excelled in single-turn alignment but struggled with long-term, multi-turn, and sparse rewards. Newer result-driven RL methods like GRPO and DAPO enhance training stability and efficiency [12]. - The evolution of algorithms involves leveraging foundational models to provide priors while refining capabilities through interaction and rewards in task environments [12]. Core Capabilities: Planning - The pipeline paradigm views planning as automated reasoning and action sequence search, which is limited in flexibility and stability under complex tasks [14][15]. - The model-native paradigm integrates planning capabilities directly into model parameters, enhancing flexibility and robustness in open environments [15][18]. Core Capabilities: Tool Usage - Early systems embedded models in fixed nodes, lacking flexibility. The model-native transition internalizes decision-making regarding tool usage, forming a multi-objective decision problem [21][22]. - Challenges remain in credit assignment and environmental noise, which can destabilize training. Modular training approaches aim to isolate execution noise and improve sample efficiency [22]. Core Capabilities: Memory - Memory capabilities have evolved from external modules to integral components of task execution, emphasizing action-oriented evidence governance [27][30]. - Short-term memory utilizes techniques like sliding windows and retrieval-augmented generation (RAG), while long-term memory focuses on external libraries and parameter-based internalization [30]. Future Directions - The trajectory of Agentic AI indicates a shift towards deeper integration between models and their environments, moving from systems designed to use intelligence to those that grow intelligence through experience and collaboration [44].

Core Insights - Cognizant Technology Solutions (CTSH) is set to report its third-quarter 2025 results on October 29, with earnings estimated at $1.29 per share, reflecting a 3.20% increase year-over-year [1][2]. Revenue Expectations - The company anticipates third-quarter 2025 revenues between $5.27 billion and $5.35 billion, indicating growth of 4.6%-6.1% and a year-over-year increase of 5.63% based on the Zacks Consensus Estimate of $5.33 billion [2][10]. Performance Drivers - Cognizant's performance is expected to benefit from an expanding clientele and a strong pipeline, including six large deals won in the second quarter of 2025, two of which were valued at $1 billion each [4][10]. - The growing demand for GenAI solutions in sectors like financial services, healthcare, and manufacturing is anticipated to provide continued growth opportunities, particularly in fraud detection and predictive maintenance [5]. - New partnerships, such as the collaboration with WRITER and the launch of Cognizant Agent Foundry, are expected to enhance enterprise-scale adoption of Agentic AI [6]. Challenges - Despite the expanding clientele, weak demand in the products and resources segment due to tariff policy concerns and spending pressures poses a challenge [7]. - Macroeconomic uncertainties and ongoing cost optimization pressures across some sectors are also expected to impact performance [7]. Strategic Partnerships - Cognizant's partnerships with major companies, including Pearson and Google Cloud, are likely to drive growth by enhancing AI-driven solutions and customer engagement [8][9][12]. - The collaboration with Pearson aims to integrate generative and agentic AI to create innovative education and workforce solutions [9]. Conclusion - Cognizant's expanding clientele and robust AI-driven solutions are expected to contribute positively to its growth prospects in the upcoming quarter [13].

Core Insights - The article discusses Marc Benioff's controversial comments regarding public safety in San Francisco and the implications for his company, Salesforce, amidst a backdrop of political and media scrutiny [1][4][5]. Group 1: Public Safety and Political Context - Benioff has been a long-time advocate for public safety and has expressed concerns about police staffing levels in San Francisco, suggesting the National Guard could assist if necessary [1][6]. - His comments were interpreted as a shift towards conservative politics, which sparked significant media backlash and public debate [4][5]. - Following the controversy, Benioff clarified his stance, stating he does not believe the National Guard is needed for safety in San Francisco and emphasized the importance of community partnership [5][6]. Group 2: Economic Impact of Dreamforce - The Dreamforce event, hosted by Salesforce, attracts nearly 50,000 attendees and is expected to generate approximately $130 million in revenue for San Francisco, creating around 35,000 local jobs [7]. - The event serves as a platform for showcasing Salesforce's innovations, particularly in the field of agentic AI, which is gaining traction among major companies [11][12]. Group 3: Salesforce's AI Innovations - Salesforce is focusing on agentic AI, which allows for autonomous decision-making and complex workflow execution without constant human intervention, distinguishing it from traditional AI models [8][11]. - The company has seen significant adoption of its agentic AI technology, with over 12,000 customers utilizing it, leading to a 120% year-over-year growth in its Data and AI business segment, reaching $1.2 billion [13]. - Benioff has articulated a long-term vision for Salesforce to become an AI-first company, with a revenue target exceeding $60 billion by fiscal year 2030, indicating a 10% organic compounded annual growth rate [15][16]. Group 4: Market Performance and Perception - Despite the potential of its AI capabilities, Salesforce's stock has declined nearly 25% year-to-date, contrasting with gains seen by competitors like Microsoft and Oracle [9]. - The market's short-term focus has often overlooked the long-term vision and transformative potential of Salesforce's innovations, as highlighted during the Dreamforce event [10][16].

This table was published 9 months ago on how Agentic AI will disrupt Financial Services: https://t.co/ooKrzcXcju ...

Core Insights - Sidetrade has completed the acquisition of ezyCollect, a leading Australian SaaS provider specializing in Order-to-Cash solutions, marking a significant step in its international growth strategy [1][2] Financial Details - The acquisition values ezyCollect at €37.3 million (A$66.5 million), with €34.7 million (A$61.9 million) paid in cash and €2.6 million (A$4.6 million) in Sidetrade shares [2] - Founders of ezyCollect may earn up to €5.6 million (A$10 million) based on the company's Annual Recurring Revenue (ARR) growth through 2028, with an estimated current earnout of €3.3 million (A$6 million) [2] - The acquisition is financed through Sidetrade's cash reserves and a €25 million long-term credit facility at an interest rate of approximately 3.1% [3] Strategic Implications - The acquisition retroactively consolidates as of October 1, 2025, and enhances Sidetrade's mission to democratize its AI platform, Aimie, aimed at improving financial performance for mid-market and large enterprises [4] - Sidetrade plans to align long-term incentives for ezyCollect's key employees through additional share grants as part of an equity plan [3] Company Overview - Sidetrade operates a SaaS platform that analyzes $7.2 trillion worth of B2B payment transactions daily, anticipating customer payment behavior and attrition risk for 40 million buyers globally [5] - The company has a global presence with over 400 employees across Europe, the United States, and Canada, serving clients in more than 85 countries [5]

Core Insights - A significant majority (71%) of banking professionals believe that investing in AI agents is essential for maintaining competitiveness, while over half (54%) express a desire for stronger regulatory guidance [1][11] - The research indicates that 70% of respondents predict agentic AI will have a "significant" impact on the banking industry, with many viewing it as a "game changer" [3][4] Current State of Agentic AI in Banking - There is a notable divide in the adoption of agentic AI based on institution size, with larger banks leading the way; 64% of larger banks are testing or actively deploying agentic AI compared to only 38% of smaller institutions [8] - Among those already deploying agentic AI, 94% report satisfaction with their current implementations [4][6] Key Use Cases - The primary use cases for agentic AI identified by banking professionals include: - Customer Service (59%) - Employee Efficiency (39%) - Fraud/Risk Management (37%) [9] - A majority (73%) believe agentic AI will enhance customer experience, and 79% think it will improve employee experience [6] Adoption Barriers and Incentives - Major concerns hindering adoption include security and compliance (51%), concerns about autonomy (51%), integration with legacy systems (46%), and costs (42%) [17] - To facilitate adoption, over half of the respondents indicated that stronger regulatory guidance (54%) and seamless integration with existing tools (53%) would be beneficial [11] Technology Perception - Following a demonstration of SoundHound's Amelia 7 platform, 88% of respondents found the technology valuable for their business, and 98% believed it represented an improvement over current technology [12] Industry Outlook - The research underscores a strong enthusiasm for agentic AI among larger institutions, with a recognition of its potential to fundamentally change the banking industry through enhanced reasoning, action, and task performance capabilities [13]

Core Insights - The convergence of financial and communication services is driven by a digital-first world, leading banks and telcos to form partnerships to leverage this trend [1][2] - Banks aim to become the central resource for all aspects of a customer's financial life, expanding beyond traditional banking services [2][3] - The concept of a financial choice hub is introduced, which centralizes various financial services and empowers consumers to make informed decisions [3][4] Group 1 - The digital convergence of services is fueled by consumer preference for accessing multiple services from a single trusted source, prompting cross-industry collaborations [1][2] - Banks can enhance customer experience by blending convenience with visibility, appealing deals, and objective insights into individual financial needs [3][4] - The financial choice hub is more than just an aggregation tool; it serves as a finance-focused superapp that proactively assists consumers in making better financial choices [4] Group 2 - The financial choice hub allows consumers to set their financial goals, link accounts, and activate options, but also introduces risks from new competitors like fintechs [4][5] - An example of the financial choice hub in action is provided, illustrating how it can help users save money by finding the best deals based on their spending habits [5]

Core Insights - The article emphasizes that agentic AI will provide value only when companies redesign workflows around outcomes rather than tasks and rethink human and AI collaboration [1][3] Group 1: Agentic AI Systems - Agentic AI is described as a "team of digital colleagues" capable of autonomous reasoning and coordination, which can transform fragmented operations into seamless workflows [3][4] - Successful implementation requires appointing "mission owners" to define goals, empowering AI agents, and structuring work around results instead of departments [3][4] - Three imperatives for building agentic AI systems include designing around outcomes, unlocking data silos through shared business logic, and developing responsible governance frameworks [4][3] Group 2: Security in Agentic Commerce - Cloudflare is collaborating with Visa and Mastercard to create a trust framework for AI agents in digital payments, ensuring secure transactions [5][6] - The proposed system uses cryptographic protocols to authenticate AI agents, allowing merchants to recognize legitimate agents without confusion [6][7] - Cloudflare plans to integrate these protocols into its Agent SDK, facilitating secure transactions while complying with payment network requirements [8][6] Group 3: Financial Institutions and Agentic AI - UBS positions agentic AI as a foundational shift in business evolution, embedding autonomous intelligence into its operations [9][10] - The bank is restructuring its operating model to manage AI agents responsibly, with IT teams overseeing autonomous systems through defined permissions and compliance [11][10] - UBS has established an Agentic AI Center of Excellence to coordinate innovation and ensure ethical deployment across various teams [12][11]

Industry Overview - The technology sector is experiencing strong growth driven by the adoption of Artificial Intelligence (AI), Generative AI (GenAI), and Agentic AI, alongside the digitalization wave fueled by AI and cloud computing [1] - There is a rising demand for high-bandwidth and low-latency networks (5G), as well as hardware components such as GPUs, AI accelerators, memory chips, and data center infrastructure [1][3] Earnings Outlook - Several technology stocks are expected to report quarterly results soon, with a focus on four companies: Meta Platforms (META), Reddit (RDDT), InterDigital (IDCC), and AMETEK (AME), which are well-positioned to exceed earnings estimates [2] - Meta Platforms anticipates third-quarter 2025 ad revenues of $48.5 billion, reflecting a 21.6% year-over-year growth, with earnings expected at $6.60 per share, indicating a 9.45% increase from the previous year [10][11] - Reddit projects revenues between $535 million and $545 million for the third quarter of 2025, suggesting a 57.8% year-over-year growth, with earnings expected at 52 cents per share, representing a 225% increase from the prior year [14] - InterDigital expects third-quarter earnings between $1.52 and $1.72 per share, with a consensus estimate of $1.79 per share, indicating a 9.8% growth year-over-year [16][17] - AMETEK forecasts mid-single-digit percentage growth in overall sales year-over-year, with adjusted earnings expected between $1.72 and $1.76 per share, reflecting a 4-6% increase from the previous year [19] Investment Trends - The demand for data center capacity is increasing due to escalating AI needs, prompting major cloud providers like Amazon, Alphabet, Microsoft, and Meta Platforms to invest significantly in AI-related infrastructure [4] - The semiconductor industry is witnessing robust growth, with semiconductor sales reaching $64.9 billion in August 2025, a 21.7% increase year-over-year [5] - The PC segment also saw growth, with IDC estimating 75.8 million units sold in the third quarter of 2025, up 9.4% year-over-year [6]

Core Insights - C3.ai's agentic AI initiative is becoming a significant growth driver, with the introduction of the C3 Agentic AI Platform and the Strategic Integrator Program aimed at enabling enterprises and government agencies to create domain-specific AI applications [1][10] - The platform addresses key generative AI challenges, ensuring enterprise-grade reliability and achieving a high success rate in large-scale deployments across various sectors [2] - C3.ai has secured early deployments with notable clients like the U.S. Army and Nucor, showcasing its potential in agentic AI adoption [3][10] - The company is targeting a $2 trillion addressable market, with increasing government adoption and OEM partnerships, positioning itself for a leading role in the agentic AI segment [4] Company Developments - The C3 Agentic AI Platform combines generative pretrained transformers with secure workflows, effectively mitigating risks such as data exfiltration and cybersecurity threats [2] - The platform's open architecture prevents vendor lock-in and is supported by 131 pre-built enterprise AI applications, enhancing its competitive differentiation [3] - C3.ai's OEM model allows partners to license and customize its technology, expanding market reach without incurring additional sales costs [10] Competitive Landscape - Competitors like Palantir Technologies and BigBear.ai are emerging as significant challengers in the agentic AI space, with Palantir expanding its AI platform for government and commercial clients [5][6] - BigBear.ai is developing mission-ready agentic AI tools, increasing competition for enterprise-scale AI transformation projects [7] Financial Performance - C3.ai's stock has seen a decline of 31% over the past three months, contrasting with a 3.5% decline in the industry [8] - The company trades at a forward price-to-sales ratio of 7.8X, significantly lower than the industry average of 17.12X, indicating potential undervaluation [12] - Analyst sentiment has become more cautious, with fiscal 2026 EPS estimates widening to a loss of $1.33, reflecting deeper expected losses compared to previous estimates [14]