Год: 2025

Retrieval augmented generation (RAG) is altering the way we use large language models, but building these systems can be hectic. In this article, you will learn how to build RAG systems using Haystack.

Learn math through roadmaps, courses, tutorials, Python frameworks for solving equations, guides, exercises, textbooks, and more.

The debate over H-1B visa has intensified ahead of US President-elect Donald Trump’s inauguration ceremony on January 20. For Indians working in their hometown and planning to move abroad, the changes in the policy have become a topic of utmost importance, particularly for the Indian IT sector.

According to a report by Macquarie Research, the ongoing challenges in hiring local talent in the US highlight the critical role of H-1B visas in addressing the talent gap. The US has a continuous shortage of qualified technical workers, further exacerbated by the 2.9% unemployment rate in November last year.

This suggests that proposed reforms to H-1B visa policies could benefit Indian IT firms. Despite their relatively low share of total visa positions – around 65,000 every year plus 20,000 for those who receive higher education from US universities – they would still be able to fill those positions.

Indian IT companies rely heavily on the H-1B visa program to deploy skilled workers in the US.

Macquarie has decided to maintain an “outperform” rating for TCS, Infosys, and HCLTech, which suggests that the firm believes Indian IT companies are well-positioned for this change.

Meanwhile, Mrinal Rai, assistant director and principal analyst at ISG, told AIM that the majority of the Indian IT providers were already preparing and reducing their dependency on the H-1B by investing in hiring more local talent.

“Given India’s vast talent pool, service providers have made significant strides in closing the skill gaps. Besides, many providers have offered upskilling programs to support local talent,” Rai said.

Rai added that global big-tech companies are impacted more than Indian IT companies. “Technology providers such as Apple, Amazon, Google and Microsoft have been more significantly impacted by the tightening of H-1B policy than the top H-1B applicants, including Cognizant, Infosys, TCS and Wipro,” he said.

Flat Wage Floor Concerns

Macquarie has raised concerns about a proposal to introduce a flat wage floor for H-1B visa holders. A flat wage is a fixed salary that doesn’t adjust for differences in job roles, locations, or cost of living. The report highlighted that such a measure might not be feasible, given the wide disparities in the cost of living across the US.

To reduce these issues, Macquarie suggested transforming H-1B visas into general work permits, similar to Norway’s skilled work permit model. This shift would allow greater job mobility for visa holders, promote competition among employers, and enhance worker protections.

If such changes are implemented, it would largely benefit the Indian IT workforce applying for visas as they would account for regional economic disparities and the complexities of the global talent market.

Krishna Vij, VP of IT hiring at TeamLease Digital, told AIM that larger IT firms are likely to adapt more effectively to the policy change due to their advanced training programs and operational flexibility. “However, smaller and mid-tier firms could benefit if the policy promotes a more equitable distribution of visas, as these companies typically submit fewer applications.”

She added that the firms that have invested in localised operations and US-based hiring will gain a competitive edge, which will reduce their dependence on H-1B visas. This includes firms like Infosys and TCS, which might navigate the changes more effectively. Their experience in managing visa allocations and compliance makes them adaptable to new regulations with minimal disruption.

On the other hand, smaller IT firms or those heavily reliant on H-1B visas may encounter greater challenges. “Mid-tier and smaller IT companies may face heightened challenges due to increased competition for visas, especially if wage-based prioritisation becomes a deciding factor,” Vij said.

She said that this will result in many smaller firms accelerating their localisation strategies, focusing on hiring US-based talent to mitigate visa risks. “The quota limit will intensify competition among Indian IT companies and force them to submit higher numbers of applications to secure the required slots.”

This might also result in companies prioritising hiring in other countries, like Canada or Mexico, for nearshore delivery centres. “These strategies will help mitigate risks, but they might also increase short-term costs,” Vij added.

Furthermore, past compliance issues and fines have prompted Indian IT firms to enhance their adherence to regulations. For example, Infosys had earlier agreed to pay millions of dollars in fines for US visa violation cases.

Already a Decline

Data from the National Foundation for American Policy (NFAP), a US-based non-partisan think tank, revealed a decline in H-1B visa approvals for Indian IT companies for FY2024. The seven largest Indian IT firms secured approval for just 7,299 new H-1B petitions – a significant drop from 14,792 approvals in 2015.

This is around a 50% decline in acceptance rate for visa approvals. Furthermore, the 7,299 approvals represented only 5.2% of all H-1B visa approvals for FY24, which equates to just 0.004% of the US civilian workforce. Denial rates for H-1B applications were low at 2.5% – slightly down from 3.5% in FY23, as per NFAP’s findings.

For example, Amazon topped the list of H-1B approvals for initial employment in FY24, with 3,871 approvals – down from 4,052 in FY23 and 6,396 in FY22. Cognizant followed with 2,837 approvals, Infosys with 2,504, and TCS with 1,452. The number accounts for less than 0.1% when we compare the number of visa applications for smaller firms.

Meanwhile, a separate report also revealed that Indian IT giants TCS, Wipro, Infosys, and HCL have reduced their reliance on the H-1B visa by 56%. These figures reflect a continued downward trend in H-1B visa approvals for Indian IT firms, but it is expected to change with the current policies or at least stay unharmed.

The post The H-1B Visa Policy Change Might be Good News for Indian IT appeared first on Analytics India Magazine.

Google and Meta have revolutionised marketing by turning data into performance analytics. Can something similar be done to ease the burden on teachers by making the process more data-driven?

This idea sparked the journey of Kushal Sinha and Piyush Kumar, the founders of Chicago-based LearnQ.ai, a smart learning platform. Their efforts culminated in VEGA (Virtual Entity for Guidance and Assistance), a specialised AI agent designed to guide and assist with any task.

It represents their effort to develop a teaching assistant that can support educators and institutions, all while maintaining the core AI and data-driven approach.

In an exclusive interview with AIM, Sinha, an IIT Guwahati alumni, said, “At this stage, AI is deeply integrated into the platform, consuming vast amounts of data to provide recommendations and insights. It thereby enhances the teaching experience and empowers educators with actionable intelligence.”

The duo wanted to take a focused approach. So, instead of going too broad, they decided to test the platform on a specific use case. “We chose the SAT exam, which is widely taken in the US for undergraduate admissions. That was our initial use case. Recently, however, we’ve started allowing early customers and clients to create their own courses,” Sinha added.

Now, they are enabling people to build courses tailored to their needs. In about a month, they’ll complete the rollout and open the platform to everyone, allowing users to build courses from scratch and deploy them.

“When we say “deploy,” the idea is to capture student interaction data—what they’re learning, where they’re struggling—and expose this data to both teachers and our AI assistant, VEGA.

“VEGA acts as an AI tutor or avatar, leveraging the student’s knowledge graph built through interactions, assessments, quizzes, and even chat history. It also integrates the professor’s expertise,” Sinha explained.

AI Assitant is Not AI Tutor

Kumar, the other co-founder and a home tutor, told AIM that unlike traditional AI tutors, which often rely solely on prompts or pre-trained large models, their approach integrates real-time student data.

“For example, think of a smartwatch. If you’re self-motivated, you’ll act on its suggestions; if not, someone—like a family member—would nudge you. Learning works similarly. Some students are self-motivated, but most need additional support.”

Sinha mentioned that they use data pipelines to build a detailed knowledge graph for each student. This enables their AI system to cater responses to the student’s specific understanding level. It’s not just a wrapper around a large language model, it’s an agentic AI system that integrates data from multiple streams to deliver personalised, actionable insights.

People are now creating courses for AP, JEE, and even unique topics like teaching the Mahabharata and Ramayana. “This highlights the flexibility of our platform. Institutions like The Doon School in India are already using it to teach K-12 students,” said Sinha. Another organisation they are actively in discussions with is Allen in India.

Why Edtechs Fail?

“The first generation of ed-tech platforms was largely driven by the rise of the internet. Platforms like Coursera and Udemy focused on accessibility. They allowed teachers to record lectures and broadcast them widely, building platforms that connected learners to content,” Sinha mentioned.

According to him, these platforms didn’t address efficiency. Even with all the advancements in ed-tech, students relied on tutoring services, marketplace-based learning models, and physical tuition centres.

These platforms were more of an add-on rather than a replacement or an efficiency enhancer in education. That’s where LearnQ.ai’s approach differs, as they aim to solve for efficiency in education.

What’s Next?

Recently, Sinha and Kumar met with American entrepreneurs working with public schools in several states, who narrated a compelling success story. Four years ago, their schools ranked last in the district.

However, by implementing a data-driven approach and using simple Excel sheets, their performance saw a sharp improvement, pulling them to the top spots in the district. Most students now score within a tight 70 to 90 range.

That explains how any small technological adoption in the education system can bring vast improvements in a student’s performance. With AI, things are looking even better.

In the current landscape, many AI tutors lack a robust data layer. A truly effective one must be able to understand where each student is in their learning journey and provide tailored guidance. This becomes specifically important since teachers often struggle to give personal attention to each student due to time constraints.

For instance, if you’re watching a lecture on Coursera and don’t understand a concept, there’s no way to directly ask the instructor for clarification. This is where AI avatars for teachers and virtual assistants that provide 24/7 access step in.

Another much-needed AI intervention would be a feature to transcend language barriers. For example, a teacher who speaks only Hindi and English can have an avatar that interacts in Korean or Mandarin. This would dramatically expand accessibility, allowing the same educator to engage with thousands of students simultaneously, addressing doubts and providing real-time support.

Another upcoming feature empowers educators and creators to design their own AI tools, which will align with their unique processes and expertise. For instance, a blogger could train the AI to replicate their specific writing style or workflow using a knowledge base they provide.

These customisations open up endless possibilities for personalisation and efficiency across educational and creative fields.

The post How These Indian Entrepreneurs Help US Students Crack SAT appeared first on Analytics India Magazine.

The year 2024 witnessed the global adoption of an agentic AI force. From big tech companies to emerging startups, AI agents took centre stage. Keeping the momentum going, 2025 is set to be even bigger in the agentic space. A noticeable trend shows that startups building AI agents are largely being founded by Indian developers.

The AI agents market has been witnessing promising growth. From $5.1 billion in 2024, the market is expected to hit $47.1 billion by 2030. In particular, Indian entrepreneurs have been significantly driving the growth.

Source: X

India and AI Agents

“If you look at how LLMs have matured, it’s clear that their potential plateaus once they run out of fresh training data. AI agents pick up the slack by plugging into external sources [such as] databases, APIs, real-time feeds, and letting the model stay relevant as events unfold,” Ramprakash Ramamoorthy, director of AI research at Zoho and ManageEngine, told AIM.

“This is why you’re seeing founders worldwide, including those from India, pushing AI agents to the forefront; it’s simply the next logical milestone for the technology after robust LLMs.”

Ramamoorthy believes that the effort to integrate AI with new data streams represents an evolution rather than a revolution and is not defined by nationality or geography but by the determination to prevent AI from becoming stagnant. “AI agents give us that infusion of timeliness and context that static LLMs alone can’t deliver.”

Bengaluru-based AI startup Kogo AI, founded by Praveer Kochhar and Raj K Gopalakrishnan, is building AI agents and solutions to simplify workflows and improve productivity for businesses. Recently, they also launched an AI agent store.

“We are currently building an agent that can look at a database and actually think like a data scientist or a business analyst and generate extremely intelligent questions,” Kochhar said in a recent podcast with AIM.

The founders also emphasised the advancements in foundational models, which have made it easier and cheaper to build AI agents.

Kochhar explained how building customer support and voice-based AI agents was initially expensive, costing around ₹50 per interaction, which was unsustainable for practical use. However, advancements significantly reduced this cost to approximately ₹2.5 per interaction, making such deployments more feasible.

A number of Indian AI agents have been making the mark. A few Indian founders under the Y Combinator cohort are also building AI agents. For example, Floworks, founded by Sudipta Biswas and Sarthak Shrivastava, selected under the Winter 2023 cohort of Y Combinator, is building AI agents that will address sales functions.

Prominent AI startups such as Sarvam AI, backed by PeakXV Partners, Khosla Ventures and CoRover, backed by Venture Catalysts, and educational institutions also have AI agents.

AI Agents are Everywhere

With AI agents quietly becoming the norm, choosing the right sector where these agents will become the most beneficial becomes critical. “Departments such as sales, marketing, and finance usually have well-established software systems like CRM (customer relationship management), ERP (enterprise resource planning), analytics dashboards, etc., so they can plug AI agents directly into these data pipelines,” Ramamoorthy said.

Industry leaders who have moved on to start their ventures have also gotten into AI agents. CP Gurnani, co-founder of AlonOS and former CEO of Tech Mahindra, recently spoke about how AI agents can make people more productive and efficient. “Agentic AI is the software version of a personal robot. Each one of us will have an AI agent that knows us really well,” he wrote on social media.

AlonOS, co-founded by Gurnani and Rahul Bhatia, provides organisations with AI-as-a-service and data engineering solutions. The Singapore-headquartered AI startup recently partnered with Indosat, Indonesia’s telecom company, to accelerate AI sovereignty in their country. Though not many details have been revealed, AI agents will probably be implemented, considering that they are first catering to the travel and hospitality sector.

Gaurav Aggarwal, who has industry experience working with NVIDIA and autonomous mobility, is now the founder and CEO of RagaAI – an AI testing platform. They recently released test frameworks for testing AI agents too. “We’re seeing AI agents evolve into so much more than just tools. They’re becoming collaborators, problem-solvers, and decision-makers,” Aggarwal said recently.

Notably, leading database company Redis is supporting AI startups such as Kore AI in powering their virtual AI agents. “The Indian tech ecosystem is going to play a very critical role in agentic AI,” Manvinder Singh, VP of AI product management at Redis, told AIM.

The post Why Indian Founders Love AI Agents appeared first on Analytics India Magazine.

The US-based semiconductor firm Synaptics announced a collaboration with Google to enhance Edge AI capabilities for Internet of Things (IoT) devices. The partnership aims to integrate Google’s machine learning (ML) core with Synaptics’ Astra hardware, accelerating the development of smart, context-aware devices.

The collaboration focuses on multimodal processing, enabling devices to seamlessly process inputs like vision, sound, and voice for applications in wearables, appliances, entertainment systems, and more. These innovations will benefit a wide range of industries, including consumer electronics, automotive, and enterprise systems.

“We are on the brink of a transformative era in Edge AI devices, where innovation in hardware and software is unlocking context-aware computing experiences that redefine user engagement,” said Vikram Gupta, senior vice president and general manager of IoT processors at Synaptics. “Our partnership with Google reflects a shared vision to leverage open frameworks as a catalyst for disruption in the Edge IoT space.”

Synaptics CEO Michael Hurlston said Semiconductor firms such as Qualcomm, AMD, and Intel, which operate in India, usually handle part of the solution in India, another part in the US, and yet another in Europe.

Synaptics’ Astra platform combines low-power silicon with open-source software, aiming to make AI solutions easier and faster to develop. Google’s ML core, designed for efficiency and compatibility with various systems, will complement Synaptics’ strengths in AI hardware.

“Synaptics’ embrace of open software and tools and proven AI hardware makes the Astra portfolio a natural fit for our ML core,” said Billy Rutledge, Director of Systems Research at Google Research. “We look forward to working together to bring our capabilities to the broad market.”

The post Synaptics, Google Join Hands to Advance Edge AI for IoT Devices appeared first on Analytics India Magazine.

Several Indian students today aspire to become AI experts—building cutting-edge models, mastering ML algorithms, and staying on top of global research trends. The desire is largely fuelled by the global AI wave and the dream being sold at Indian schools and universities, which are all introducing AI and ML in their curricula.

AIM Research recently published a report on the top 50 institutes offering AI courses in India 2025, highlighting how universities are increasingly introducing courses for AI in master’s programmes. This is driven by the increasing industry demand and corporate partnerships for providing students with real-world applications of AI/ML and upskilling them with the relevant skills for the industry.

However, according to several industry experts, the curriculum of Indian universities offering these courses in their regular computer science fields seems out of place. There is a big mismatch between the curriculum and what the industry demands.

How Much is Too Much?

BITS Pilani campus group vice-chancellor V Ramgopal Rao told AIM that while AI is being introduced in many universities, it often lacks depth. Topics like deep learning, reinforcement learning, and advanced AI applications are either optional or lightly covered.

“There is limited focus on hands-on projects or exposure to real-world problems, which are essential for AI development,” Rao said. “AI is a fast-evolving field, and the lack of emphasis on research and innovation within the curriculum limits students’ ability to contribute to global advancements.”

He emphasised that AI requires knowledge of fields like cognitive science, ethics, and domain-specific expertise (e.g., healthcare, finance), which are rarely integrated into CS programs. “Universities need to partner with industries to include hands-on AI projects, internships, and exposure to real-world applications within the existing framework,” he said.

Similar sentiments were expressed by Chiranjib Sur, a visiting professor of computer science at Krea University in Andhra Pradesh, who told AIM that the current curriculum on AI in most universities focuses on using Python libraries to build AI models. That needs to change.

“I think that we need to calibrate the courses and teach more foundational things,” Sur said, citing the importance of teaching the science and mathematics behind those libraries and how to build them. This will ensure that the students and the professionals know which techniques to use for different problems.

Sur said that a stronger emphasis on fundamental mathematics in combination with core knowledge of computer science (systems, architecture, compilers etc) is essential.

“While I teach computer science or some AI techniques, the focus is to learn those techniques and solve domain-specific problems,” he added, highlighting the need for learning fundamentals of CS like mathematics and not just AI applications.

Ritvik G, a graduate research assistant at the Artificial Intelligence Institute of South Carolina (AIISC), told AIM that the curriculum followed across Indian universities places a heavy emphasis on classroom-based learning and theoretical exams, with limited attention to practical lab work or real-world projects.

“AI-related courses generally focus on traditional topics like NLP, computer vision, and algorithms. However, foundational architectures for advanced AI workloads, such as neural architecture search, enterprise AI or neuromorphic computing, are rarely part of the curriculum,” he said.

Ritvik further pointed out that the education system was set up to churn out developers and individuals focused on getting jobs rather than adapting to the evolving needs. “Four exams that centred around a book and few slides rather than intuitive thinking, where anything outside of their answer key is regarded as the wrong answer with no score provided for anything close but perfect,” he added.

There is definitely a shift needed in the CS degrees in Indian universities. “Teach math, statistics, programming fundamentals, and soft skills in the first year. The second year should be all about various projects using GenAI tools. Problem-solving is more important than coding,” said Naveed Iftikh, co-founder of AtomCamp.

Should Indian Universities Offer a Different Course for AI?

“I’ve noticed many colleges introducing separate ‘AI & ML’ branches, like an exclusive department apart from Computer Science. The syllabus? 80% the same as CS, with just some buzzwords like ‘Neural Networks’ and ‘TensorFlow’ added,” Chaitanya Athukuri, product manager at Nielsen, said in a post on LinkedIn.

He added that though AI and CS are not intrinsically separate fields, the education providers are able to sell it well.

Meanwhile, many AI experts recommend that people study computer science and take a crash course in generative AI, if necessary.

There are differing voices too. Amit Sheth, the chair and founding director of AIISC, told AIM that he believes that AI has advanced and expanded so much in this century that it is no longer ideal to have it as a part of the CS curriculum.

“In my opinion, it is necessary to offer degrees and skills-based certifications in AI, in addition to CS. In a four-year program, it may be appropriate to have one or two years in common, but there is so much to learn at foundational, translational, and application levels in AI,” Sheth said.

Priyankar Kumar, a software engineer, said that the curriculum is largely the same worldwide—the problem is that it focuses on exams instead of practical applications.

AIM spoke to two recent CS graduates from the RNS Institute of Technology in Bengaluru, who agreed that the new AI curriculum was not up to date with the current industry requirements. They said that though there is a separate AI-ML branch in BTech, the curriculum often overlaps. At the same time, they believe that the two separate branches are necessary as the topics differ starting the second year of college.

Similarly, Adarsh Shirawalmath, the founder of Tensoic, had earlier told AIM that though his college has been helpful, it is still years behind when it comes to the type of research that is being pursued there.

According to several online forums, people want to study AI, ML, and data science in their BTech classes but always find the curriculum outdated. Moreover, several students also question the idea of AI as a separate discipline, as the course is very similar to the CS major.

The post Indian Universities Desperately Need to Update Their Outdated CS Curriculum appeared first on Analytics India Magazine.