StartupSuperb

Tag: ml

  • Apple’s Ambitious Strategy to Launch a Compelling Cloud Platform and Compete with AWS and Google Cloud

    Apple’s Ambitious Strategy to Launch a Compelling Cloud Platform and Compete with AWS and Google Cloud

    “`html

    Apple’s Developer-Focused Cloud Initiative: Project ACDC

    Apple is reportedly evaluating the launch of a cloud service tailored for developers, leveraging its proprietary silicon. This move could place Apple in direct competition with established players such as Amazon Web Services (AWS), Microsoft Azure, and Google Cloud. A comprehensive investigation by The Information’s Aaron Tilley reveals that this concept, internally known as Project ACDC (Apple Chips in Data Centres), was a significant topic of discussion within Apple throughout the first half of 2024.

    The Cloud Market: Expansive, Profitable, and Competitive

    Cloud infrastructure services form the backbone of the modern internet. As per Amazon’s earnings report, AWS alone generated $25 billion in revenue during Q1 2024. Meanwhile, Microsoft Azure and Google Cloud are also witnessing rapid growth, with Microsoft’s Intelligent Cloud segment achieving $26.7 billion in the same quarter.

    These platforms empower businesses and developers to deploy applications, store data, and execute compute-heavy tasks such as AI training and inference without the need to manage physical servers.

    Understanding Apple’s Project ACDC

    As highlighted by MacRumors in 2024, Apple investigated the possibility of allowing developers to rent servers powered by its M-series chips, which are utilised in its Macs and iPads. These chips are designed on the Arm architecture, recognised for their energy efficiency and AI inference capabilities.

    Apple has already integrated its own silicon in several data centres, particularly for its Private Cloud Compute infrastructure, which serves as a secure server-side backbone for its recently unveiled Apple Intelligence features. This system processes user data with privacy in mind while utilising Apple’s Neural Engine for large-scale inference.

    The report also mentions that Apple tested these Mac chip-powered servers with the Siri team focusing on text-to-speech functionalities, before broadening their application to services like Apple Music, Photos, and Wallet.

    Why Is Apple Contemplating This Market Entry?

    1. Services Revenue Faces Scrutiny

    Apple’s Services division generated $85.2 billion in FY2023, according to its annual report. However, increased regulatory scrutiny regarding App Store commissions in the EU and the United States, coupled with ongoing antitrust investigations concerning its $20 billion annual Google search agreement, pressurises Apple to broaden its revenue sources. A developer-centric cloud platform could serve as a new growth opportunity.

    2. Maximising the Silicon Edge

    Apple’s chip division stands out as one of its critical technological differentiators. If Apple can provide comparable or superior performance for cloud-based AI inference at lower energy and hardware expenditures, it presents a significant value proposition, especially amidst growing AI adoption.

    3. Managing the Complete Stack

    Apple’s strategy has consistently centred around controlling hardware, software, and services. A cloud offering powered by Apple Silicon would enhance the company’s end-to-end control over performance, privacy, and the developer experience.

    What Could Apple’s Cloud Look Like?

    No official announcement has been made, but based on current reports, a developer-centric Apple Cloud might offer:

    • Seamless integration with tools such as Xcode, Swift, and Core ML
    • Facilitation of deployment for on-device machine learning models on a large scale
    • A more private, Apple-friendly alternative to third-party services like AWS
    • Potential inclusion under the iCloud umbrella, albeit with a dedicated developer segment

    Notably, Apple reportedly did not intend to establish a traditional enterprise sales division. Instead, its Developer Relations team was earmarked to oversee access, creating a more intuitive onboarding process compared to AWS or Azure.

    Technical Focus: AI Inference Over Training

    Apple’s M-series chips and Neural Engine are crafted for on-device inference, which entails executing pre-trained AI models to deliver real-time predictions or classifications. While entities such as OpenAI and Meta train extensive models using NVIDIA H100s or Google TPUs, the actual deployment of these models for applications like voice processing, object recognition, or smart replies can be efficiently managed by more optimised hardware.

    Apple employs this technology for features including:

    • Visual Look Up in Photos
    • Siri’s on-device responses
    • Language translation
    • Personalisation in Apple Music and News

    This makes Apple Silicon exceptionally suitable for cloud-based inference tasks, especially for applications that do not demand the vast scale or versatility of AWS’s entire GPU stack.

    Will Apple Proceed with This Initiative?

    The future of this initiative remains uncertain. Michael Abbott, Apple’s former VP of Cloud Engineering and a pivotal supporter of Project ACDC, departed the company in 2023. Although discussions reportedly extended into early 2024, the current state of the project is unclear.

    Nonetheless, the internal use of Apple Silicon in its own infrastructure indicates that even if Apple opts not to release this as a public-facing service, it will continue to utilise this capability to enhance its offerings.

    “`

  • Meesho Unveils Key Features of BharatMLStack for Enhanced AI Integration

    Meesho Unveils Key Features of BharatMLStack for Enhanced AI Integration



    Meesho’s E-commerce AI Solutions with BharatMLStack

    Meesho’s E-commerce AI Solutions with BharatMLStack

    E-commerce platform Meesho has taken a notable step by open-sourcing essential components of its internal machine learning platform, BharatMLStack, aimed at enhancing access to AI infrastructure for Indian startups and developers.

    Key Features of BharatMLStack

    The open-source release, now available on GitHub, features Meesho’s feature store, orchestration interface, control plane, and software development kits. Developed over a span of two years, BharatMLStack meets the company’s substantial daily data processing requirements of approximately 1.91 petabytes and facilitates real-time applications.

    Impressive Performance Statistics

    In FY25, Meesho’s machine learning systems recorded an astonishing 66.9 trillion feature retrievals and generated 3.12 trillion real-time predictions. These advanced systems were instrumental during key events such as the Mega Blockbuster Sale that took place in March 2025.

    Support for AI Development

    By making its feature store available, Meesho seeks to assist data teams in developing AI systems tailored for real-time applications like fraud detection and recommendation engines. This feature store ensures reliable access to precomputed data signals, significantly reducing the time needed to deploy machine learning models.

    Future Plans for BharatMLStack

    Meesho has plans to introduce more components of BharatMLStack in future updates. The platform is now open for contributions from developers via GitHub.

    Strategic Business Developments

    In addition to this technological advancement, the company successfully completed its reverse flip to India after obtaining final approval from the National Company Law Tribunal last month.


  • Unveiling the Future: Google Pixel 10 Series Set to Launch on August 13 with Enhanced AI, Camera, and Performance

    Unveiling the Future: Google Pixel 10 Series Set to Launch on August 13 with Enhanced AI, Camera, and Performance



    Pixel 10 Series Launch: All You Need to Know

    Pixel 10 Series Launch

    Google is set to unveil the Pixel 10 series on 13 August, marking a return to the mid-August launch window initially established last year with the Pixel 9 series. Multiple sources, including the trustworthy leaker @MysteryLupin and a report from Android Headlines, confirm that the event is confirmed, with pre-orders anticipated to open the same day and shipments commencing on 20 August, which is two days before last year’s schedule.

    This announcement closely follows an exclusive Pixel Superfans event held in London, where Google invited select guests to preview the upcoming devices. Although there was speculation that this event might signal an earlier release, it now appears that Google will adhere to its August timeline.

    Four Models, Familiar Design, and a Powerful New Chip

    The upcoming Pixel lineup is expected to feature four distinct models: the Pixel 10, Pixel 10 Pro, Pixel 10 Pro XL, and the Pixel 10 Pro Fold. The overall design is anticipated to remain largely consistent, featuring flat edges, a horizontal camera bar, and slim bezels reminiscent of the Pixel 9 Pro. However, significant upgrades are projected under the surface.

    The entire series is believed to be equipped with Google’s next-generation Tensor G5 chip, reportedly developed by TSMC. This advance could enhance performance, boost power efficiency, and address longstanding heating and battery issues noted in prior models.

    Android 16 with On-Device AI Features

    On the software side, the Pixel 10 series is expected to launch with Android 16, which is set to incorporate deeper integration of AI features. New functionalities might include Video Generative ML for footage editing, Speak-to-Tweak voice commands for Google Photos, and Sketch-to-Image capabilities, allowing rough sketches to be transformed into realistic images, all executed on-device to ensure speed and privacy.

    Camera Enhancements and Strategic Downgrades

    One of the more unexpected developments may involve the introduction of a triple-camera setup on the base Pixel 10, a feature that was previously reserved for the Pro models. However, some leaks also indicate that Google might use older sensors from the Pixel 9a, which could include a downgrade in the ultra-wide lens. Nonetheless, Google’s renowned computational photography is likely to sustain its reputation for strong image quality.

    What to Expect

    With just over two months remaining, leaks suggest that the Pixel 10 series will not represent a dramatic overhaul. Instead, it is likely to offer a refined update focused on AI features, enhanced performance, and incremental hardware upgrades. The continuity in design and user experience appears to be part of the strategy, while meaningful improvements beneath the surface aim to maintain the competitiveness of the Pixel lineup.

    Stay tuned to Startup Superb for more updates as the official unveiling approaches.


  • Unpacking Neysa’s Series A Funding: Valuation Insights and Ownership Structure

    Unpacking Neysa’s Series A Funding: Valuation Insights and Ownership Structure

    “`html

    Neysa Secures $30 Million in Series A Funding

    Generative AI startup Neysa has successfully raised $30 million in its Series A funding round. This round was led by NTT Venture Capital, with participation from Nexus Venture Partners, Z47 (previously known as Matrix Partners India), Anchorage Capital, and Rajesh Kumar Dugar. The announcement, made in October of the previous year, saw contributions from the company’s founder and CEO Sharad Sanghi, along with CPO Karan Kirpalani.

    Neysa’s board has granted 1,11,532 compulsory convertible preference shares from Series A at an issue price of Rs 20,430. Additionally, the company has issued 12,592 equity shares, collectively raising Rs 252 crore or $30 million, as per its regulatory filing with the Registrar of Companies (RoC).

    In this round, NTT Ventures led the funding with an investment of Rs 75.67 crore or $8.9 million. Both Nexus Venture Partners and Z47 (formerly Matrix Partners India) contributed Rs 67.26 crore ($7.9 million) each. Anchorage Capital added Rs 16.8 crore to the mix, while Sharad Sanghi allocated Rs 25.22 crore in equity shares. Other individual investors contributed to the remaining amount.

    This new funding will enable Neysa to enhance its AI infrastructure, boost research and development efforts, and prepare for the introduction of its Gen AI acceleration cloud service, according to the company.

    Recently, Neysa issued equity shares worth Rs 20 crore to founder Sharad Sanghi and Anchorage Capital. The shareholding and valuation figures are based on data following this latest issuance.

    It is noteworthy that the conversion terms for Series A shares will be 1:2. According to estimates from Startup Superb, the company’s valuation post-allotment is expected to reach around Rs 1,090 crore or $128 million.

    Neysa’s Offerings in Generative AI

    Neysa is focused on providing AI cloud solutions and platform-as-a-service (PaaS) for users to effectively manage and scale Generative AI projects. Its product lineup includes Nebula, a platform for deploying and scaling AI workloads; Palvera, which enhances network intelligence; and Aegis, designed to secure AI/ML ecosystems against new and emerging threats.

    Based in Mumbai, Neysa has raised a total of $50 million so far, which includes a $20 million seed round. Data from various startup intelligence platforms indicates that Nexus Venture Partners and Z47 each hold a 16.22% stake, making them the largest external shareholders in the company. NTT VC and Anchorage Capital hold stakes of 14.83% and 4.32%, respectively. The founders, Sharad Sanghi and Aninya Das, collectively maintain a 43.09% share of the company.

    Financial Performance

    As of FY24, Neysa has not reported any operating revenue. However, the company did accrue interest income amounting to Rs 13.87 lakhs through fixed deposits with banks and security deposits during the fiscal year ending March 2024. In the same timeframe, Neysa recorded a net loss of Rs 3.1 crore.

    “`

  • Microsoft and Yotta Join Forces to Propel India’s AI Revolution

    Microsoft and Yotta Join Forces to Propel India’s AI Revolution



    Accelerating AI Adoption in India through Microsoft and Yotta’s Partnership

    Accelerating AI Adoption in India through Microsoft and Yotta’s Partnership

    Accelerating AI adoption in India is now a reality thanks to the partnership between Microsoft and Yotta Data Services. This collaboration will enable Microsoft to incorporate its Azure AI services with Yotta’s Shakti Cloud platform, thereby providing AI capabilities to a diverse audience that includes developers, startups, enterprises, and public sector organisations.

    Integration of Technology for Enhanced AI Capabilities

    The agreement merges Microsoft’s AI models, applications, and development tools with Yotta’s sovereign cloud infrastructure, aimed at delivering high-performance AI computing solutions within India. This infrastructure will facilitate low-latency model training and real-time inferencing across various key sectors such as agriculture, healthcare, education, finance, manufacturing, retail, and media.

    Supporting the IndiaAI Mission

    This partnership is in line with the objectives of the IndiaAI Mission, which, as of May 2025, has attracted over 500 proposals for developing indigenous AI models. Microsoft and Yotta intend to engage with government agencies, academic institutions, and startups to foster local innovation and assist in the creation of AI models tailored for India’s digital public infrastructure.

    Azure AI Services Available on Yotta’s Infrastructure

    Microsoft’s Azure AI platform, featuring tools like ML Studio, database services, security offerings, and GitHub, will become accessible on Yotta’s GPU infrastructure. This collaboration aims to facilitate the deployment of hybrid AI systems that comply with the country’s data sovereignty requirements, while maintaining high standards of performance, scalability, and safety.

    Promoting India’s AI Ambitions

    Puneet Chandok, President of Microsoft India and South Asia, pointed out that the partnership aligns with India’s ambition to become an AI-first nation. He mentioned that Microsoft is privileged to contribute to the country’s AI goals through innovations that address India’s specific needs.

    Access to Language Models for Yotta Customers

    Customers of Yotta using the Shakti Cloud will have the opportunity to access a variety of foundational and specialised language models (LLMs and SLMs) via the Azure AI Foundry. This platform includes integrated safeguards such as content filters, groundedness detection, and copyright protection to assist organisations in scaling their AI initiatives responsibly.

    Transforming AI Accessibility in India

    Sunil Gupta, Co-founder, CEO, and Managing Director at Yotta Data Services, regarded the collaboration as a significant leap towards India’s self-reliance in AI and digital transformation. He highlighted that this partnership would enhance the accessibility of advanced AI capabilities to businesses of all sizes within India.

    This announcement further complements Microsoft’s previous commitments to AI development in India. In January 2025, Microsoft Chairman and CEO Satya Nadella disclosed a partnership with IndiaAI, a division of the Digital India Corporation, which introduced the establishment of an AI Centre of Excellence and AI Productivity Labs aimed at promoting inclusive growth.


  • Google Pixel 10 Unveiled: Sneak Peeks at Design, Tagline, and AI Innovations Before the Big Reveal

    Google Pixel 10 Unveiled: Sneak Peeks at Design, Tagline, and AI Innovations Before the Big Reveal

    “`html

    Google Pixel 10: Anticipated Smartphone Release with AI Innovations

    Google seems to be intensifying preparations for the introduction of its upcoming flagship smartphone, the Pixel 10. Recently, new leaks have surfaced, allegedly from a commercial filming session, showcasing insights into the phone’s design while hinting at a variety of AI-driven features that are set to characterise this new device.

    Discovered by a user on the platform X (previously known as Twitter), the images from behind the scenes display a professional film crew engaged in what is being described as a comprehensive commercial shoot for the Pixel 10. This shoot involved sophisticated equipment, including a macro probe lens and a Panavision rig, with over 20 crew members dedicated to capturing footage of an individual simply holding the smartphone. The detailed storyboards featuring the labels “Google” and “Pixel 10” lend further authenticity to the claim, as the visual style corresponds with Google’s established marketing approach.

    Just out for a walk… stumbled onto a full-on commercial shoot for the Google Pixel 10. They had a macro probe lens, a Panavision rig, and 20+ crew members… to film someone holding a phone. If the Pixel camera’s so good, why not just use it? #BTS #Vancouver pic.twitter.com/muDluZfK75

    — Mark Teasdale ★ (@MarksGonePublic) May 23, 2025

    Marketing Strategy and AI Features

    This campaign seems to revolve around the tagline: “Ask more of your phone,” likely referencing the new array of AI functionalities Google highlighted during its I/O 2025 developer event. If the messaging is indicative, the Pixel 10 is set to place a significant emphasis on generative AI, potentially making it the most AI-centric smartphone that Google has released.

    Design and Key Features of the Pixel 10

    In terms of aesthetics, the Pixel 10 appears to closely mirror the design of the Pixel 9 Pro. Leaked images exhibit a black “Obsidian” variant of the device, showcasing a triple-camera setup that was previously exclusive to the Pro model, indicating that this feature may now be available across the entire Pixel 10 series. Elements such as the power button and volume controls remain in their familiar locations. Additionally, the storyboards reveal two circular sensors located near the camera array, likely housing a flash and a temperature sensor, resembling the design of the Pixel 9 Pro.

    Notably, the storyboard suggests new camera functionalities, with one panel dedicated to a feature tagged “Add Me,” which may point towards an AI-enhanced capability for group photography or image compositing. Another anticipated feature titled “Video Generative ML” is expected to debut in Google Photos, offering AI-facilitated video editing options. This is complemented by “Speak-to-Tweak,” a voice-controlled photo editing tool, and “Sketch-to-Image,” which can transform simple user sketches into realistic images, akin to Samsung’s Galaxy AI.

    Hardware and Software Anticipations

    Previous leaks indicate that the Pixel 10 will be powered by the next-generation Tensor G5 chip, with 12GB of RAM as standard. The Pixel 10 Pro may elevate this further with 16GB of RAM, staying consistent with the Pixel 9 Pro series. A Geekbench listing has already shown that the Tensor G5 is currently being tested on Android 15, though it is expected that the Pixel 10 will be released with Android 16, representing the OS’s first appearance.

    In terms of photography, conceptual renders suggest a robust triple-lens system featuring a 64MP main sensor, a 64MP ultra-wide lens, and a 64MP periscope telephoto lens, potentially providing up to 50x digital zoom. These specifications stem from speculative artistic interpretations and ought to be regarded with caution until officially confirmed by Google.

    Expected Launch Timeline for the Pixel 10

    Although Google has not officially confirmed the Pixel 10’s release, the company generally unveils its flagship smartphones during late summer. With last year’s Pixel 9 series debuting in August, a similar timeline is expected for the Pixel 10.

    As this commercial shoot leak stands as the most credible pre-launch revelation to date, expectations are soaring that the Pixel 10 will enhance Google’s hardware strategy and significantly elevate the standard for AI integration in smartphones.


    “`

  • Perfios Expands Its Horizons with IHX Acquisition to Innovate Health Insurance Technology

    Perfios Expands Its Horizons with IHX Acquisition to Innovate Health Insurance Technology

    “`html

    B2B SaaS Fintech Firm Perfios Expands with Acquisition of IHX

    B2B SaaS fintech company Perfios is making strides by acquiring healthcare information exchange IHX, aiming to strategically enhance its technology offerings within the health insurance claims sector. This acquisition follows a successful funding of approximately $2.58 million raised by IHX from LogX Venture Partners and Quik Solutions.

    This acquisition marks the third significant purchase for the Bengaluru-based entity in a mere three months, as it seeks to strengthen its expertise across the banking, financial services, and insurance (BFSI) landscape.

    Perfios plans to merge its data intelligence and analytics capabilities with the IHX platform to enhance the speed of claim processing, improve operational visibility, and facilitate real-time decision-making for hospitals, insurers, and patients alike. This will be bolstered by leveraging its existing network of 30,000 hospitals and 30 insurers.

    As part of the acquisition arrangement, IHX will maintain its independent operations, brand identity, and leadership structure, mirroring the approach taken with Perfios’ previous acquisitions.

    About IHX

    Established in 2020 by Mayank Tiwari, Ajay Bakshi, Sanjay Kalra, Ambuj Kashyap, and Mahesh Nagaraj, IHX functions as a healthcare data exchange entity. It operates a digital platform that connects various stakeholders within the healthcare system, including hospitals, payers, and diagnostic laboratories. The focus remains on enabling data exchange, managing consent, and offering actionable insights driven by AI/ML technologies.

    IHX claims to manage over 40% of India’s cashless health insurance claims, processing more than 10 million transactions annually, equating to a substantial claim value of $1 billion. Its platform links 30,000 hospitals across 1,200 locations with over 30 insurers, featuring prominent names such as Reliance Hospital, DY Patil Hospital, MGM Healthcare, and Jaslok Hospital.

    Perfios’ Recent Acquisition History

    In the previous year, Perfios expanded its portfolio by acquiring Karza Technologies and FintechLabs Technologies. Shortly after, it acqui-hired Chennai-based Fego.ai, a startup focusing on behavioural financial insights. The acquisition of IHX is part of a series of strategic moves, which also includes the earlier acquisitions of Clari5, a banking fraud detection platform, and CreditNirvana, an AI-enabled debt collections and recovery solution, both from 2025.

    Global Expansion and Financial Growth

    Currently, Perfios operates in 18 countries, with ongoing efforts to broaden its international presence. The company’s revenues demonstrated robust growth, increasing by 37% year-on-year to reach Rs 558 crore in FY24, up from Rs 407 crore in FY23.

    “`

  • OpenAI Unveils Essential Handbook for Developing LLM Agents in Real-World Scenarios

    OpenAI Unveils Essential Handbook for Developing LLM Agents in Real-World Scenarios

    A Practical Guide to Building Autonomous AI Agents

    OpenAI has introduced a comprehensive and technically sound guide titled A Practical Guide to Building Agents, which is specifically designed for engineering and product teams interested in the development of autonomous AI systems. By leveraging real-world applications, the guide provides a systematic method for recognising appropriate use cases, designing agents, and implementing robust safeguards to ensure their reliability and safety.

    Understanding an Agent

    The term agent refers to autonomous systems that carry out multi-step tasks with minimal human intervention, unlike traditional LLM-powered applications like single-turn chatbots or classification models. These advanced systems incorporate reasoning, memory, tool functionality, and workflow administration.

    An agent consists of three core elements:

    • Model — The LLM that drives decision-making and reasoning.
    • Tools — External APIs or functions utilised to execute actions.
    • Instructions — Structured prompts outlining the agent’s goals, behaviours, and limitations.

    When to Consider Developing an Agent

    Developing an agent is advantageous for workflows that surpass the limits of standard rule-based automation. Common scenarios include:

    • Complex decision-making tasks, such as intricate refund approvals in customer service.
    • Rule systems that require constant maintenance, like compliance workflows that are fragile or challenging to scale.
    • Engagement with unstructured data, including document parsing and contextual natural language communication.

    The guide stresses the importance of thorough validation to confirm that an agent-level reasoning process is necessary before implementation begins.

    Technical Essentials and Overview of the SDK

    The OpenAI Agents SDK presents a versatile code-first framework for creating agents using Python. Developers can define agents declaratively by choosing models, registering tools, and specifying prompt logic.

    OpenAI categorises tools into the following groups:

    • Data tools — For retrieving context from databases or document stores.
    • Action tools — For writing or updating data, and triggering downstream services.
    • Orchestration tools — Agents themselves that can be called as sub-modules.

    Instructions should stem from operational procedures and be articulated in clear, modular prompts. The guide suggests the use of prompt templates with variable parameters for increased scalability and maintainability.

    Orchestration Approaches

    The guide discusses two main architectural models:

    • Single-agent systems — A single, looped agent managing the complete workflow, ideal for simpler scenarios.
    • Multi-agent systems:
      • Manager pattern — A central coordinator assigns tasks to specialised agents.
      • Decentralised pattern — Peer agents independently exchange control among themselves.

    Both designs support adaptable execution paths while maintaining modularity via function-based orchestration.

    Implementing Safeguards for Safe and Predictable Operations

    The guide details a multi-layered protective strategy to reduce risks, including data leakage, inappropriate responses, and potential system abuse:

    • LLM-based classifiers — For ensuring relevance, safety, and detecting personally identifiable information (PII).
    • Rules-based filters — Including regex patterns, input length constraints, and enforcement of blacklists.
    • Tool risk assessments — Assigning sensitivity ratings to external functions and controlling their execution accordingly.
    • Output verification — Making sure responses are consistent with the organisation’s tone and compliance standards.

    Guardrails are built into the agent runtime, allowing simultaneous assessment and intervention when violations occur.

    Incorporating Human Oversight and Escalation Mechanisms

    Recognising that even well-crafted agents might face ambiguity or need to handle critical actions, the guide advocates for the inclusion of human-in-the-loop strategies. These strategies include:

    • Failure thresholds — Escalating issues after multiple misinterpretations or tool failures.
    • High-stakes actions — Directing irreversible or sensitive tasks to human oversight.

    Such approaches facilitate gradual implementation and enable trust to develop over time.

    The guide positions OpenAI’s framework as a method for designing intelligent agents that are functional, manageable, and ready for production. By amalgamating sophisticated models with targeted tools, structured prompts, and stringent safeguards, development teams can transition from experimental prototypes to resilient automation infrastructures.

    The framework serves as a strong foundation for integrating agents into practical applications, whether managing customer workflows, processing documents, or generating developer tools. OpenAI advises initiating with single-agent deployments and gradually expanding to multi-agent orchestration as complexity increases.

  • Unlocking Potential: MGID’s CTR Guard Boosts Viewable Click-Through Rates by 29%

    Unlocking Potential: MGID’s CTR Guard Boosts Viewable Click-Through Rates by 29%



    CTR Guard: The Innovative AI-Powered Tool from MGID

    CTR Guard: The Innovative AI-Powered Tool from MGID

    CTR Guard is MGID’s new AI-driven solution aimed at tackling ad fatigue and boosting digital ad efficiency. By employing machine learning (ML) and generative AI (GenAI), CTR Guard anticipates and mitigates drops in clickthrough rates (CTR), helping advertisers maximise their campaign effectiveness with less manual input.

    According to 41% of marketers, ad fatigue is one of their primary concerns. When consumers repeatedly see the same advertisements, the effectiveness of campaigns diminishes quickly, with CTR decreasing by an average of 15% within just a week of launching a campaign. Conventional A/B testing techniques typically provide a reactive solution to ad fatigue, often failing to address the issue until performance has already declined.

    In response to this widespread challenge, MGID has created CTR Guard, an advanced tool that continuously tracks campaign performance, forecasts CTR reductions before they happen, and automatically proposes AI-generated creatives to keep audience engagement intact.

    Key Features of CTR Guard

    CTR Guard stands out from traditional ad optimisation tools by combining advanced predictive analytics with AI-driven creative development to maintain campaign freshness and interest. The process is highly efficient and impactful:

    Early Detection

    CTRG Guard’s proprietary ML algorithms scrutinise both historical and real-time campaign data to identify early indicators of ad fatigue.

    Automated Alerts

    Advertisers will receive immediate email notifications and dashboard alerts if their CTR is predicted to fall below a crucial threshold.

    AI-Powered Creative Optimisation

    CTR Guard suggests three AI-generated ad creatives, featuring images and headlines tailored to achieve advertising goals.

    One-Click Deployment

    Advertisers have the option to approve, adjust, or reject AI-generated creatives, ensuring they maintain complete control over campaign execution while reducing manual workload.

    Early assessments show that CTR Guard’s predictive model has over 90% accuracy in spotting potential CTR drops before they significantly affect performance. This allows CTR Guard to enhance the viewable CTR (vCTR) by an average of 29% by aligning AI-generated creatives with current audience engagement trends.

    Kenneth López Triquell, VP of Sales at MGID, remarked that advertisers constantly strive to keep their audiences engaged, and traditional strategies against ad fatigue are often too slow and inefficient. He believes CTR Guard signifies a significant shift in ad optimisation, automating laborious, data-heavy processes while giving advertisers full control over their creative direction.

    Although AI is central to MGID’s innovative solution, human oversight remains crucial. Unlike opaque AI systems, CTR Guard ensures that advertisers maintain complete control over decision-making, offering a transparent and user-centric approach to enhancing ad performance through AI.

    CTR Guard is now accessible to MGID advertisers globally. For further details, please visit MGID’s blog.

    Press Contact

    For all enquiries, please reach out to: hello@mgidpr.com

    About MGID

    MGID operates as a global advertising platform that supports brands and publishers in succeeding on the open web through innovative, AI-enhanced native marketing. By employing privacy-conscious, AI-based technology, MGID delivers high-quality, relevant advertisements in secure environments, reaching over 1 billion unique visitors each month. The company’s varied advertising formats, which include native, display, and video options, effectively balance user experience and performance, raising brand awareness while allowing publishers to profitably engage their audiences.

    Headquartered in Santa Monica, MGID maintains a global presence with 18 offices worldwide. The company’s commitment to technology, talent, and strategic partnerships has sustained a five-year track record of double-digit growth year-on-year. As MGID broadens its influence in North and South America, Europe, and Asia, it remains focused on sustainable and profitable growth, continuously adapting its products to assist both sides of the supply chain in navigating the ever-evolving landscape of digital advertising.


  • “Unlocking Opportunities: Sabeer Bhatia Champions Stanford’s Innovative Learning Approach and Sparks IIT Curriculum Discussion”

    “Unlocking Opportunities: Sabeer Bhatia Champions Stanford’s Innovative Learning Approach and Sparks IIT Curriculum Discussion”


    IIT Curricula Debate Ignited by Hotmail Founder Sabeer Bhatia

    Sabeer Bhatia, who is well-known as the founder of Hotmail, which is now part of Microsoft Outlook, has initiated a discussion surrounding the relevance of the curriculum at Indian Institutes of Technology (IITs). He emphasised that while Stanford University provides current knowledge, much of what is taught at IIT remains outdated.

    Critique of IITs’ Outdated Curriculum

    The concerns regarding the curriculum at IITs being obsolete have been ongoing. Critics often point out that the strong focus on theoretical knowledge does not adequately prepare graduates for contemporary job roles that require hands-on experience in areas such as artificial intelligence, data science, blockchain, and other emerging technologies. The slow modifications to the curriculum, along with inflexible course structures, further inhibit the integration of current industry trends and interdisciplinary learning opportunities.

    Sabeer Bhatia’s Perspective on Learning

    In a recent post on X (formerly Twitter), Bhatia stated that Stanford’s approach teaches students what is relevant today, while IITs remain rooted in the past. He shared his own experience, noting that although his grades helped him secure a position at Apple, it was his on-the-job learning that allowed him to establish Hotmail. Bhatia remarked that the true source of education today comes from freely available knowledge online, insisting that innovation stems from practical experience rather than mere study.

    Reactions from the Online Community

    Bhatia’s comments struck a chord with many. One individual observed that, despite ongoing initiatives to incorporate practical learning into educational programmes, there exists considerable resistance. They noted their experience teaching a practical innovation course at IIM-A for a decade in collaboration with a corporate partner. This course emphasises real-world applications with minimal theory, and it has gained significant popularity among students. However, there are still some faculty members, particularly those deeply entrenched in traditional academia, who frequently criticise such initiatives.

    Another commenter highlighted the discrepancy in employment data, stating that a startling 36% of IIT Bombay’s latest graduating class could not secure job placements in a challenging job market. This situation underscores the immediate necessity to update IIT curricula in accordance with industry needs and to tackle the broader issue of unemployment within India’s educated youth.

    Diverse Opinions on the Current State of IITs

    Despite the criticism, not everyone aligns with Bhatia’s viewpoint. One user respectfully disagreed, asserting that his information appears outdated. They recalled that even in 2011, professors at IIT Kanpur were teaching cutting-edge courses, particularly in artificial intelligence and machine learning, covering essential topics up to training personal neural networks. This individual highlighted the importance of industry partnerships and project-based learning at IITs, suggesting that Stanford’s advantages stem more from its connections to Silicon Valley than from any deficiency in innovation at IIT institutions.

    Another participant brought a philosophical perspective to the discussion, suggesting that Stanford has a unique advantage due to its ability to stay connected to contemporary developments, prioritising real-time relevance over traditional rote learning. They acknowledged IITs’ rigorous standards but indicated that these can sometimes lead to the entrenchment of outdated academic frameworks. Bhatia’s own journey, which transitioned from academic success to practical achievement, serves as evidence that free knowledge widely available on the internet acts as a powerful equaliser. While degrees may open doors, it is through practical experience that true success is built.

    The Ongoing Need for IIT Curriculum Adaptation

    Bhatia’s insights have revitalised discussions surrounding the necessity for IITs to revisit and reorganise their curricula. This is essential for maintaining relevance in a rapidly evolving technological environment.