A naukri.com initiative
Data Science News
Analyticsindiamag
1M
122
Image Credit: Analyticsindiamag
Microsoft Launches New Security Updates in Ignite 2024
- Microsoft announced several security updates at the Ignite 2024 event.
- The Security Exposure Management tool is now available, providing a unified view of enterprise security posture.
- Microsoft launched a $4 million bug bounty called Zero Day Quest for investigating cloud and AI vulnerabilities.
- Updates and new skills were announced for Copilot AI, including risk analysis and support for security operations.
Read Full Article
7 Likes
Medium
1M
40
Image Credit: Medium
Python is No Longer the King of Data Science
- New programming languages such as Julia and R are emerging and catering to the needs of modern data science, chipping away at Python's dominance. Julia is built for speed from the ground up and is particularly appealing for heavy computations requiring high performance, optimization and scientific simulations. R still holds its ground as the first choice for statisticians and academic researchers for advanced statistical modeling. Specialized tools such as SQL and industry-specific platforms like SAS and MATLAB offer solutions tailored to specific needs.
- Python is slow compared to many of its rivals and struggles with performance-intensive tasks. Python's inherent inefficiencies are being highlighted by the need for libraries such as NumPy and Pandas that leverage optimized backend code. Python's Global Interpreter Lock limits the language's ability to perform true multi-threading, making it harder to harness the power of modern multi-core processors.
- For projects requiring large-scale parallel computations, Python's limitations can become a bottleneck. The sheer number of libraries available within Python can confuse users as to which to choose. Big data technologies like Apache Spark and Hadoop designed for distributed computing are better suited for handling efficiency in organizations that deal with terabytes of data.
- Python's reliance on external frameworks such as TensorFlow and PyTorch for deep learning has exposed Python's weaknesses in handling complex computations directly, and some developers are exploring alternatives such as Julia or C++. Real-time analytics is another area where Python falls short and other languages such as Java and platforms like Apache Kafka are better suited for low-latency operations.
- Despite these challenges, Python remains the easiest and most efficient choice for many data science tasks. However, the future of data science will be polyglot, and data scientists will increasingly use multiple languages and tools depending on the task at hand. Julia could become the go-to choice for high-performance computations, while R will continue to excel in advanced statistics and visualization.
- SQL will always have a role in database management, and emerging platforms like DuckDB may further streamline SQL-based analytics. Lower-level languages such as C++ or Java may gain more ground for deep learning and production-level systems.
- Python's role is shifting from being the all-encompassing "king" of data science to being an important piece of a much larger puzzle. Organizations and data scientists who embrace this diversity will be better equipped to tackle modern challenges by combining the strengths of multiple languages and tools to unlock new possibilities and push the boundaries of what's possible in data science.
- Python's decline as the "king" of data science isn't a fall from grace but an evolution. Data science has outgrown the need for a single ruler, and a more diverse ecosystem means more innovation, better performance, and ultimately, better outcomes for everyone involved.
Read Full Article
2 Likes
Medium
1M
446
Image Credit: Medium
The future of software development
- Artificial intelligence (AI) is revolutionizing software development, automating tasks and enhancing testing processes.
- Low-code and no-code platforms are democratizing software development, enabling non-technical users to create applications.
- Remote work has introduced collaboration tools, fostering diverse development teams and breaking down geographical barriers.
- Developers are prioritizing cybersecurity integration throughout the software development lifecycle.
Read Full Article
26 Likes
Medium
1M
346
Image Credit: Medium
Science and Technology Common people consider AI-produced poetry better than humans
- A study by the University of Pittsburgh found that ordinary people consider AI-produced poetry to be written by humans.
- The research showed that AI technology was 75% more likely to recognize poems written by real poets than those generated by AI itself.
- Ordinary people rated AI-generated poems as higher quality than the work of real poets, potentially due to their simplicity and common sense.
- Researchers noted that while AI can mimic the creative work of famous poets, it lacks the depth of emotion, surprise, empathy, and other aspects that make poetry truly meaningful.
Read Full Article
20 Likes
Analyticsindiamag
1M
154
Image Credit: Analyticsindiamag
No One Gets Named and Shamed Like Indian IT Hiring
- An Indian IT hiring and onboarding practice involving delays has come under fire, with developer forums and social media platforms full of complaints.
- Various IT companies have been identified for the practice of offering letters to freshers then delaying onboarding, while some rescind offers citing potential cost-cutting with rising AI adoption being a key factor.
- On IT forum Reddit, a developer complained that no joining letters were being given to freshers in LTIMindtree 2024 batch, with only a handful being accepted as interns and no joining dates given.
- Companies like Cognizant and TCS have also been accused of the practice, with many graduates placed in support roles rather than developer positions.
- According to latest earnings carry out by Indian IT giants such as Infosys, Tech Mahindra, and Wipro, freshers will be onboarded within 15,000 to 20,000 headcount in FY25, promising to reduce their bench size and hire new graduates.
- However, there are still concerns that recruiting efforts will be minimal following the onboarding debacle of 2023, with potential layoffs being a threat.
- Aiming to resolve the consistent hiring issues, developers opine that candidates should apply multiple times to various IT companies while keeping an eye on other job opportunities.
- This problem has existed for years, with applicants waiting for months to receive responses to their applications and interviews.
- The Indian IT industry needs to overhaul its hiring process to prevent the talent exodus.
Read Full Article
9 Likes
Medium
1M
259
Image Credit: Medium
The Invention of the Computer
- The concept of mechanical computation dates back to ancient civilizations, with the development of the abacus by the Sumerians in 2300 BCE.
- During the 17th century, mechanical machines such as the Pascaline and the Step Reckoner were invented, paving the way for modern computers.
- The Pascaline, invented by Blaise Pascal in 1642, was capable of adding and subtracting numbers.
- The Step Reckoner, invented by Gottfried Wilhelm Leibniz in 1673, could perform addition, subtraction, multiplication, and division.
Read Full Article
15 Likes
VentureBeat
1M
259
Image Credit: VentureBeat
Microsoft’s new AI agents support 1,800 models (and counting)
- Microsoft has claimed users can now build their own custom autonomous agents or deploy out-of-the-box bots.
- Microsoft 365 Copilot now provides access to the 1,800-plus models in the Azure AI catalog.
- IDC reports that over the next two years, more and more companies will build custom, tailored AI tools.
- New integrations with Azure AI Foundry will support custom search indices and models, allowing users to access a variety of models and fine-tune them.
- Already-built models can be useful across enterprises, as Microsoft is releasing several ready-made agents that can handle simple, repetitive tasks or more complex multi-step processes.
- A new Azure AI Foundry SDK offers a simplified coding experience and toolchain for developers to customize, test, deploy and manage agents.
- McKinsey & Company is working with Microsoft on an agent that will speed up client onboarding. Meanwhile, Thomson Reuters has built an agent to help make the legal due diligence process more efficient.
- Agents are increasingly authoring processes and workflows and working across groups of people and in multi-agent systems.
- Agents serve as a layer on top of large language models, generating recommendations for humans or, if programmed, act on their own.
- Conversation data that was once siloed is now being used by agentic AI to provide context in real time.
Read Full Article
15 Likes
Dev
1M
195
Image Credit: Dev
How AI Smart Search Can Streamline Volunteer Onboarding Process?
- Nonprofits use volunteers to run their programs and make an impact in their communities. The onboarding process to bring in new volunteers can be complex as it involves a range of resources, forms, and contact information, which can be overwhelming for new recruits. AI smart search solutions help simplify the onboarding experience such that volunteers can find the resources they need efficiently.
- AI-powered smart search solutions use machine learning (ML) and natural language processing (NLP) to understand the intent of the user and provide the most relevant information.
- Smart recommendations are part of AI-powered search features that can give volunteers context on what they are searching for. This helps in offering more support to the new volunteers, guiding them intuitively through the onboarding process.
- AI smart search can search through FAQ sections and natural language questions can be used instead of scrolling endlessly through pages of information. It simplifies the experience of finding answers to more specific inquiries.
- AI-powered smart search solutions can enhance the time-consuming paperwork process by autofilling the necessary forms or showing appropriate documents that volunteers need to fill out.
- Tailored resources can be recommended to volunteers according to their specific role or interests, which makes them feel empowered and more prepared for their roles.
- Self-service options help reduce the workload of staff by enabling volunteers to find answers and resources independently. This helps volunteers find help instantly, but staff is also freed up to focus on more significant responsibilities.
- AI-powered smart search is transforming the volunteer onboarding experience such that nonprofit organizations' websites become accessible, efficient, and friendly. Implementing smart search is an impactful way to improve volunteer retention, enhance volunteer onboarding, and in effect help nonprofits fulfill their mission more effectively.
Read Full Article
11 Likes
Analyticsindiamag
1M
1.1k
Image Credit: Analyticsindiamag
Bengaluru-based KOGO Launches World’s First AI Agent Store
- Bengaluru-based AI company KOGO has launched the world's first AI Agent Store.
- The store offers businesses access to hundreds of AI tools, agents, and plugins.
- It provides ready-made agent templates, easy deployment, seamless integration, and analytics.
- KOGO's AI Store democratises AI adoption across industries with a pay-per-use model.
Read Full Article
12 Likes
Medium
1M
113
Image Credit: Medium
How Companies Are Using Closed AI Systems to Safeguard Data Sovereignty and Stay Ahead in the…
- Organizations are turning to closed AI environments to safeguard data sovereignty while utilizing AI.
- Closed AI environments are self-contained systems that operate within an organization's infrastructure.
- Options for closed AI environments include open-source LLMs, private cloud solutions, and containerization.
- Implementing closed AI environments requires investment in infrastructure and expertise, but offers enhanced data security and regulatory compliance.
Read Full Article
6 Likes
Medium
1M
415
Image Credit: Medium
The Incredible Rise of AI Agents
- AI agents are transforming various facets of business and everyday life, with sophisticated algorithms and machine learning enabling an interaction that feels almost human.
- These agents are reshaping customer interfaces, optimizing how calls are managed, and boosting company reputations in the process.
- Tools like WordLift are harnessing entity extraction and AI-driven analyses to fill content gaps and consequently boost search rankings.
- AI agents like OneAI’s OneAgent smoothly sift through content, customizing it to align with individual preferences, generating comprehensive news reports with unrivaled precision.
- AI agents require an understanding of human nuances and the most successful implementation strategies strike a balance between machine efficiency and human emotions.
- Offering a direct line to human customer service representatives remains crucial for more complex inquiries, building trust and confidence in AI-assisted services.
- AI agents unlock potential far beyond traditional methods and enhances the human component of businesses rather than replacing it.
- AI agents have increased customer satisfaction ratings, enhanced SEO performance metrics, reduced operational costs, and improved brand loyalty.
- AI agents have crafted new pathways for interaction and learning, reshaping how companies engage with customers and the digital world at large.
- The amazing potential of AI agents presents an exciting journey and together, we'll explore what lies ahead.
Read Full Article
24 Likes
Analyticsindiamag
1M
328
Image Credit: Analyticsindiamag
Goodbye Vanilla RAG, Agentic RAG is Here
- Vanilla RAG approach is reaching its limitations, where as agentic RAG is here aligning its future with AI platforms. It helps in delivering adaptable, proactive AI systems using external knowledge sources. Core principle of RAG is combined with flexibility of AI agents and assist in creating dynamic frameworks. It adapts the retrieval strategy as per the user needs and provides proactive and adaptive nature.
- Agentic RAG core objective is to integrate AI agents into the retrieval-augmented generation pipeline. These agents allow for autonomous frameworks, considering multi-step reasoning, planning and tool Utilization. It addresses issues like predefined knowledge sources, validating retrieved data and not providing iterative refinement. Various stages of retrieval are integrated with agent-based systems to provide better adaptive strategies.
- Agentic RAG integrates agents to orchestrate complex tasks involving flexible knowledge validation, departmental data retrieval, access to exclusive APIs/Databases, and multi-document retrieval. Agents are specialized in summarizing internal documents, retrieving public data, or analyzing personal content like chat logs.
- With agentic RAG, it seems like fine-tuning and RAG conversation is finally over. It provides accuracy and speed in resolution by retrieving information from internal knowledge base, documentation, and community forums. The agentic RAG framework combines LLMM layers to reason over inputs and post-process outputs.
- Agentic RAG framework is not limited to single agent systems only, multiple agents collaborate under the guidance of a meta-agent. RAG systems are needed when core AI systems are not accurate enough for information. Google introduced a new approach to move beyond RAG, retrieval interleaved generation (RIG) using LLMs with Data Commons.
- In agentic RAG, agents access varied knowledge sources beyond just databases, anticipate user needs and take preemptive actions, enabling a smoother and more efficient interaction process. It is therefore considered effective in scenarios requiring detailed reasoning, multi document comparison, and comprehensive decision-making.
Read Full Article
19 Likes
Analyticsindiamag
1M
68
Image Credit: Analyticsindiamag
AGI Won’t Happen Without Test-Time Training
- MIT has achieved a record 61.9% accuracy on the abstraction and reasoning corpus (ARC) benchmark using test-time training.
- François Chollet, creator of Keras, built the ARC-AGI benchmark to measure progress on logical reasoning abilities.
- The current leader, MindsAI, scored 55% by using a technique that fine-tunes the model at the time of testing.
- Despite MIT scoring 62%, MindsAI remains the leader due to time limit requirements and private data usage guidelines.
- MIT trained the parameters using low-rank adaptation (LoRa) and initial fine-tuning on a publicly available ARC-AGI dataset.
- The test-time training technique strengthens the model’s understanding of the ARC problem dataset by ommitting examples and learning from the rest.
- Based on the frequency of predictions, the model votes for a top prediction, evaluates the list of top predictions across transformations, retrieves the accurate output and transforms it back to the original input style.
- Test-time methods could play a pivotal role in advancing the next generation of Large Language Models.
- ARC-AGI is still the only benchmark designed to resist memorisation and measure progress to close the gap between current AI and AGI.
- As the data corpus grows, the boundaries between specialised and general-purpose models tend to blur.
Read Full Article
4 Likes
Analyticsindiamag
1M
346
Image Credit: Analyticsindiamag
Yotta Unveils Drishticam: AI Smart Surveillance Platform
- Yotta Data Services has launched Drishticam, an AI cloud-based surveillance platform designed to enhance security and operational efficiency.
- Drishticam provides real-time threat detection, facial recognition, biometric access control, and predictive analytics.
- The platform offers insights into traffic flow and peak usage hours for optimizing business operations.
- Yotta's Shambho Accelerator Programme aims to foster innovation among AI startups.
Read Full Article
20 Likes
Analyticsindiamag
1M
177
Image Credit: Analyticsindiamag
AI is the New BI
- ThoughtSpot, a California-based BI and analytics company, has launched Spotter, an agentic AI tool that functions as a virtual analyst to assist businesses.
- Spotter enables conversational interactions with data, allowing users to pose complex, multi-step questions in natural language and obtain accurate responses.
- ThoughtSpot aims to redefine self-service BI by addressing the challenges of traditional self-service BI, and providing an analytics platform built ground up for AI.
- ThoughtSpot's vision is to enable users to understand data through a conversation with their data, positioning itself as the 'Google plus ChatGPT for data.'
Read Full Article
10 Likes
For uninterrupted reading, download the app