Год: 2025

Expedia is doubling down on AI-powered travel, and CTO Ramana Thumu says the timing couldn’t be better. After stints at Fanatics and eBay, he joined the company to tap what he calls a rare intersection of technology, data, commerce and global-scale travel.

In an exclusive interview with AIM, he said, “There are very few global-scale companies with this depth of data.”

Expedia Group, an online travel agency (OTA), which operates major brands such as Expedia, Hotels.com, Vrbo, Orbitz, Travelocity and Hotwire, reported a 9% YoY revenue increase in Q3 2025. Thumu said the priority now was using AI to elevate customer experience across its massive marketplace of airlines, hotels and car rentals—a moat powered by a unified data platform.

3-Bucket AI strategy

The company has a three-part AI strategy. First is traveller-facing innovation like Trip Matching—a feature that allows users to turn Instagram Reels into real, bookable travel itineraries, along with a conversational booking feature in Hotels.com. Next comes powering hotel and advertiser partners with smarter insights, followed by boosting internal productivity through a GenAI playground used by 6,000 employees who have built 1,500 agents.

Expedia uses models from OpenAI, Anthropic and Google and is also integrated with ChatGPT, allowing users to discover trip options and refine preferences.

According to Thumu, coding assistants are translating into real productivity gains across teams. “85-88% of engineers use coding assistants. We are seeing 15-30% improvement in cycle time. That’s a real unlock,” he added.

Thumu mentioned that roughly 100 engineers across Expedia Group contributed to agentic AI initiatives worldwide, working with global product leads. Expedia also has AI squads inside departments like finance, legal and market management to automate routine tasks and make everyday work easier and faster for teams.

In customer support, more than half the issues are handled by Expedia Group’s virtual agents, CFO Scott Schenkel had said previously. The company is also using AI to generate quick summaries for human agents, helping lower the cost of each service interaction.

Not Afraid of Google

Google is introducing new agentic features in its AI Search mode, allowing it to book restaurants, events, etc., which sent Expedia’s shares down. Users simply have to state their preferences, and the AI scans multiple reservation platforms to present real-time options.

Thumu, however, remains unaffected. “Inspiration can happen anywhere,” he said. “Our job is to be at the forefront of integration. And once travellers come to us, loyalty and personalisation keep them in our ecosystem.”

He explained that Expedia’s approach is to deliver shared platform capabilities for both B2C and B2B, and then add partner-specific integrations wherever required. “We have different offerings for our B2B players—white-label, templates, Rapid API and various products,” he said.

“The beauty is that the same platform that powers B2C also powers B2B for the large part, and then we build specialised integrations so business development can move much faster and independently.”

Expedia Group competes with a wide range of major online travel platforms, including Booking.com under Booking Holdings, Airbnb, Tripadvisor, Priceline, Kayak, Trivago, Hotels.com, Skyscanner and Travelocity, among others. Many of them are already adopting AI to improve user experiences and smoothen operations.

When asked what differentiates Expedia from the others, Thumu said it was the travel data. “We flow all the data of different brands into the same data lake.”

Expedia’s Bengaluru and Gurugram hubs house 1,800 engineers and scientists driving its insurance tech, ad tech and AI platforms, with active hiring across IITs and NITs.

On skills, Thumu said, Expedia is hiring aggressively in India for mobile engineering, cloud engineering and, most importantly, deep AI skills. “AI, mobile engineering, cloud and core platform application engineering is always going to be there, but we are doubling down on AI machine learning footprint,” he said.

Expedia runs its entire insurance technology operation, including product engineering, AI and machine learning, out of its Indian offices. A large share of the company’s multibillion-dollar advertising technology stack is also developed in the country.

The post Expedia Isn’t Losing Sleep Over Google’s AI Push appeared first on Analytics India Magazine.

Two years ago, no one could have imagined that Google would suddenly leap ahead of OpenAI in the AI race.

The search giant has come a long way since Bard’s rocky debut in 2023. In its inaugural demo, the chatbot incorrectly stated that the James Webb Space Telescope had taken the first-ever image of an exoplanet, even though the first such image was captured in 2004. This mistake proved embarrassing for Google.

Following that, in 2024, Google’s image generation model, Gemini, faced criticism for producing historically inaccurate and racially biased visuals.

Between 2022 and 2024, OpenAI surged ahead as ChatGPT became a household name. Feeling the pressure, Google launched Gemini in December 2023 and tweaked its benchmark methodology to assert an edge over GPT-4.

However, with the release of Gemini 3 and Nano Banana Pro, the search giant has finally demonstrated its technical strength—so much so that its market cap is now edging towards $4 trillion, joining NVIDIA, Microsoft and Apple in the club.

Google’s stock price has been up nearly 20% since last month, compared to NVIDIA and SoftBank, which were down 7% and 35% respectively over the same period.

Google said Gemini 3 Pro outperforms OpenAI GPT-5.1 and Claude Sonnet 4.5 across significant independent AI benchmarks, including LMArena, Humanity’s Last Exam, GPQA Diamond and MathArena Apex. These benchmarks measure how effectively LLMs handle complex, human-level tasks that test their reasoning, problem-solving and real-world capability.

Since the launch of these models, social platforms have been filled with infographics and images generated by Nano Banana Pro, with users experimenting across artistic styles.

For instance, OpenAI co-founder Andrej Karpathy said he asked Gemini 3 to design a personalised workout schedule along with accompanying posters he could print and hang on the wall as reminders.

Meanwhile, Google DeepMind CEO Demis Hassabis credits the company’s strength to how research, engineering, and infrastructure teams work together. In a post on X he asserted that the company’s “real secret” is this deep integration, driven by relentless focus and intensity.

But Google isn’t alone in the race. Anthropic remains close beside, most recently launching Claude Opus 4.5, which the company claims beats Gemini 3 Pro in coding and agentic tasks in key benchmarks.

While Anthropic is doubling down on coding excellence, Google is pushing ahead on multimodality and ecosystem integration.

Soumith Chintala, co-creator of PyTorch, said on X that the launch of Gemini 3 feels “closer to the GPT-4 moment than any other in recent times”, describing the sudden burst of progress, especially with Nano Banana, as “overwhelming.”

He added that although Google now looks invulnerable with Gemini 3 backed by Tensor Processing Units (TPUs), Android and Chrome, the race is far from over.

Chintala also pointed out that Anthropic continues to dominate coding tasks, suggesting that real-world usage will ultimately determine the winner.

His comments capture a divide among users. While some prioritise strong coding performance, where Anthropic leads, others are drawn to Google’s multimodal features.

Vikrant Patankar, founding filmmaker at Composio, told AIM that Gemini 3 feels like the first time Google shipped a model family that is both powerful and practical. “The quality is noticeably stable across text, images and real-time tasks, and the Nano Banana efficiency jump makes on-device multimodal feel real instead of theoretical,” he said.

That reaction sets the tone for how the broader industry is responding. For several leaders, Gemini 3 stands out not just for performance but for how usable it feels. Salesforce CEO Marc Benioff said he was stunned by Gemini 3’s capabilities, calling the leap in reasoning, speed, images and video “insane”. Having used ChatGPT “every day for three years”, he revealed that after spending just two hours with Gemini 3, he decided he’s “not going back”.

OpenAI in Crisis?

The positive response of Gemini 3 even prompted OpenAI CEO Sam Altman to congratulate Google publicly. “Looks like a great model,” he posted on X.

However, in a recent internal memo accessed by The Information, Altman acknowledged that Google’s recent AI progress could create some temporary economic headwinds for OpenAI. “Google has been doing excellent work recently in every aspect,” he said in a compliment to the tech giant, mentioning that OpenAI is catching up fast.

“It sucks that we have to do so many hard things at the same time—the best research lab, the best AI infrastructure company, and the best AI platform/product company—but such is our lot in life. And I wouldn’t trade positions with any other company,” Altman wrote.
Despite these challenges, OpenAI continues to lead in user adoption. ChatGPT has around 800 million weekly active users as of late 2025, Altman revealed during his keynote address at OpenAI DevDay 2025.

Meanwhile, Google CEO Sundar Pichai announced during the earnings call that the company reported over 650 million monthly active users of Gemini by Q3 2025.

In a blog post, Google said that 65% of its Cloud customers are already using its AI products, a category that includes the Gemini Enterprise offering. Meanwhile, OpenAI has announced that it has surpassed one million business customers globally.

According to Patankar, Google’s new stack finally feels like a unified AI layer. Search, Android and Workspace all gain tightly integrated capabilities, from richer reasoning to real-time multimodal understanding.

He said that while OpenAI still dominates the story around a single powerful app,

Google is trying to build AI into the way people use their devices every day. “If they can keep this stable at scale, it changes the battlefield completely.”

Shrivastava said many people believed “Google Search was finished”, but AI Mode shows it is “far from over and has actually got better.” He also pointed out that more than 13 million developers are now using Google’s generative AI models.

Notably, AI Search Mode is now the default when users type into the Google search bar, with AI Overviews also surfacing automatically as results.

Sergey is Back

Part of Google’s resurgence can be linked to co-founder Sergey Brin’s return. In an interview earlier this year, Brin recalled how, at a party, an OpenAI employee named Dan encouraged him to rejoin the company. “What are you doing? This is the greatest transformative moment in computer science,” Dan had asked him. Brin returned to active work at Google in 2023 to focus on developing AI products, particularly Gemini.

Meanwhile, Pichai, in May this year, said, “I think Sergey is definitely spending time with the Gemini team in a pretty hardcore way, setting and coding and spending time with the engineers.” He added that this involvement brings unmatched momentum to the group.

Google’s NVIDIA Alternative

Besides LLMs and the software ecosystem, Google also holds an advantage in hardware. Gemini 3 was trained on Google TPUs, while OpenAI is currently building compute partnerships with Oracle, Amazon, and NVIDIA. Google’s TPUs are engineered to excel at inference workloads by offering high throughput, low latency and power-efficient compute.

According to Paras Chopra of Lossfunk, Google could “eat NVIDIA’s lunch”. He explained that Google should capitalise on the fact that Gemini 3 was trained entirely on TPUs and build on that advantage by expanding JAX and cutting TPU costs on its cloud platform.

Notably, Anthropic recently announced plans to expand its use of Google Cloud services, including the deployment of one million TPUs. Recent reports also suggest that Meta is considering adopting them.

This shift prompted a slide in NVIDIA’s stock, leading the company to issue a statement saying, “NVIDIA is a generation ahead of the industry—it’s the only platform that runs every AI model and does it everywhere computing is done.”

Currently, Google’s specialised chips (TPUs) work very well with JAX, the newer software framework that has largely taken over from TensorFlow. However, they don’t work as well with PyTorch, the most popular framework in the tech industry.

At the same time, OpenAI has partnered with Broadcom and Foxconn to build AI chips and accelerators, as well as develop data-centre networking technologies.

Tech analyst Beth Kindig captured the market’s shift in sentiment towards Google. In a post on X, she said that just nine months ago, investors believed Google was “toast” because ChatGPT’s rise threatened its dominance in search. Today, she said, sentiment has flipped so dramatically that the market now thinks Google is strong enough to challenge NVIDIA in custom silicon.

What’s Next for OpenAI?

In response to Google’s rapid advancements, Altman has reportedly told employees that OpenAI will close the gap. However, the company’s latest model, GPT-5.1, has so far struggled to capture users’ interest.

Reports further stated that the company is working on a new language model internally referred to as ‘Shallotpeat’. The model attempts to tackle the issues that surfaced during pre-training.
Besides that, OpenAI recently launched GPT-5.1-Codex-Max, a new agentic coding model that can run for over 24 hours.

The post Google Went After OpenAI But Ended up Rattling NVIDIA appeared first on Analytics India Magazine.

SoftBank Group has completed the acquisition of all equity interests in Ampere Computing. The deal makes Ampere a wholly owned subsidiary of SoftBank and brings the semiconductor firm into SoftBank’s consolidated financial statements.

SoftBank carried out the takeover through its subsidiary Silver Bands 6 (US) Corp. The company said it is reviewing the financial impact of the transaction and will disclose details if required.

Ampere, a Santa Clara based semiconductor design company, focuses on AI compute built on the ARM platform. This comes just after the Japanese giant sold all its NVIDIA shares earlier this month.

SoftBank had first announced its plan to acquire Ampere on March 20 for $6.5 billion. In that statement, SoftBank said the transaction would make Ampere “an indirect, wholly owned subsidiary” once closed and would support its broader AI strategy.

SoftBank had also noted that Ampere could complement the chip design work of Arm.

The deal was subject to US antitrust clearance, approval by the foreign investment committee and other closing conditions.

Ampere was founded in 2017 by Renée J James, its chairman and CEO. The company designs processors for cloud computing and AI workloads. Its earlier financials in the attached document show revenue of $151.8 million in 2022, $46.7 million in 2023 and a further decline the following year, along with continued losses.

SoftBank said in the new update that Ampere’s results will now be consolidated and that it will provide further disclosures if any arise from the review of the financial impact. The company also said, “Should any matters requiring disclosure arise in the future, SBG will announce them promptly.”

The post SoftBank Completes Ampere Acquisition Amid Stock Decline Tensions appeared first on Analytics India Magazine.

According to global talent firm Randstad’s Workmonitor report on AI & Equity, the vast majority of workers who say they’re skilled in AI are men at 71%, while women with AI skills stand at 29%, indicating a 42 percentage point gender gap.

Upskilling women, especially those from non-metro regions, does more than diversify the talent pool. It broadens the innovation lens, bringing in perspectives that tech teams often overlook.

Addressing this imbalance is Nimaya, a non-profit organisation quietly equipping women with the skills, confidence, and leadership capabilities required to excel in AI-driven industries.

In a podcast with AIM, Nimaya co-founders Navya Nanda and Samyak Chakrabarty talked about their vision to bridge the gender gap in technology through structured programs that combine technical training, mentorship, and real-world exposure.

“Our goal is to create an environment where women are not just participants in AI but are leading innovation and decision-making,” said Nanda. “We believe that diverse perspectives are essential to designing AI solutions that are ethical, inclusive, and impactful,” she added.

Building Skills Beyond Coding

While many tech programs focus solely on coding and algorithmic skills, Nimaya takes a broader approach. Its curriculum integrates AI fundamentals with hands-on projects, data analytics, machine learning, and ethical AI practices.

Women enrolled in the program work on real-life AI applications, ranging from predictive analytics to natural language processing, allowing them to experience the end-to-end process of AI development.

“We wanted to go beyond just teaching technical skills. Leadership, problem-solving, and strategic thinking are equally important,” said Chakrabarty.

One of Nimaya’s distinguishing features is its mentorship program. Participants are paired with industry leaders, including AI researchers, data scientists, and executives, who provide guidance on career paths, technical challenges, and personal development. This mentorship helps participants envision themselves in high-impact roles, and gain the confidence to pursue ambitious goals.

“Having a mentor who believes in you can completely change the trajectory of your career,” said Chakrabarty, adding that their focus is on inculcating leadership, beyond imparting AI training.

Mentorship also extends to collaborative projects where women work in teams to solve real-world AI problems.

Creating Opportunities in the AI Ecosystem

Nimaya actively partners with tech companies, startups, and research institutions to create internship and placement opportunities for its participants. These collaborations ensure that women not only acquire knowledge, but also gain meaningful industry experience. By integrating classroom learning with professional exposure, Nimaya helps participants transition smoothly into AI careers.

“We want our participants to leave the program with not just skills, but opportunities to lead,” Nanda said. “This approach ensures that women can make tangible contributions to AI innovation from day one.”

The program has already seen success stories across various sectors. From fintech to healthcare AI, Nimaya-trained women have contributed to building algorithms, designing ethical AI frameworks, and leading AI-powered product initiatives.

Nimaya’s work is not limited to training; it also actively advocates for gender diversity in AI. Through workshops, webinars, and conferences, the initiative highlights the importance of including women in AI decision-making roles. It also engages with policymakers and industry leaders to promote inclusive hiring practices.

“Diversity is not just a moral imperative, it’s an innovation imperative,” Nanda emphasised.

She further added that studies have consistently shown that diverse teams outperform homogeneous ones, particularly in complex problem-solving scenarios such as AI design and implementation.

Talking to AIM on the subject, Arppna Mehra, VP human resource, Honeywell India, mentioned, “India’s next big leap in AI won’t come only from flagship tech hubs. It’s already taking shape in smaller cities where young women are picking up real, industry-ready skills.”

For Nimaya, plans are underway to expand its programs across multiple cities, and to partner with global tech firms to offer specialised tracks in emerging AI domains such as generative AI, autonomous systems, and ethical AI governance.

The initiative aims to create a pipeline of women who are ready to take on high-impact roles and drive innovation in AI on a global scale.

With the right partnerships to scale this momentum, India has the potential to unlock a new generation of AI builders who won’t merely fill roles, but redefine the boundaries of what the industry believes is possible.

“The future of AI must be inclusive. Women should not be the exception; they should be the norm,” Nanda concluded. “Through Nimaya, we are committed to ensuring that women are at the forefront of AI leadership.”

The post How Nimaya is Preparing Women to Lead in AI appeared first on Analytics India Magazine.

Ilya Sutskever, former OpenAI co-founder and now CEO of Safe Superintelligence Inc. (SSI), revealed new details about the company’s strategy and eventual business model during a rare podcast appearance with Dwarakesh Patel.

Despite SSI raising nearly $3 billion since 2024 and still having no product, Sutskever insisted that monetisation will only follow after the core research is solved.

“Right now, we just focus on the research, and then the answer to this question will reveal itself,” he said. “I think there will be lots of possible answers.”

SSI began with the philosophy of “straight shot superintelligence”: prioritise building a safe superintelligent system first, and worry about productisation later. But Sutskever now says even that approach may require exposure to the public before reaching the finish line.

Patel asked why SSI would attempt to build superintelligence directly, while competitors release “weaker and weaker intelligence” models that gradually acclimate the public.

Sutskever agreed that releasing powerful AI incrementally may be necessary because society cannot meaningfully grasp its impact from essays and forecasts alone.

“It is nice to say, we’ll insulate ourselves from all this and just focus on the research and come out only when we are ready and not before,” he said.

But, he added: “I think on this point, even in the straight shot scenario, you would still do a gradual release of it… The gradualism would be an inherent component of any plan. It’s just a question of what is the first thing that you get out of the door.”

On the question of compute spending, Sutskever, who has been a vocal critic of the idea that simply “throwing more and more compute” will get the industry to superintelligence, also argued that SSI does not need the same scale of hardware investment as today’s AI giants.

He claimed that most of the enormous budgets at labs like OpenAI and Anthropic are tied up in inference, multimodal systems, staffing and product engineering — not in pure research.

“Then when you look at what’s actually left for research, the difference becomes a lot smaller,” he said. “The other thing is — that if you are doing something different —do you really need the absolute maximal scale to prove it? I don’t think it’s true at all.”

Patel challenged him further: if SSI is exploring “50 different ideas,” how can it know which ones will rival a breakthrough like the transformer without massive compute?

Sutskever responded: “I think that in our case, we have sufficient compute to convince ourselves and anyone else that what we’re doing is correct.”

Sutskever stressed that SSI’s edge is simply that it is pursuing a different paradigm, not trying to beat other labs on product cycles. “We are squarely [an] age of research company,” he said, and the goal is to first validate new ideas about generalisation before deciding how to deploy them.

Sutskever co-founded OpenAI in 2015 after leaving Google Brain, where he contributed to the breakthroughs that led to large-scale deep learning systems.

He exited OpenAI in May 2024 amid concerns that commercial pressure was overtaking the company’s original safety-first mission.

On June 19, 2024, he launched SSI with former Apple AI lead Daniel Gross and former OpenAI researcher Daniel Levy.

Operating between Palo Alto and Tel Aviv, SSI secured $1 billion by September 2024 and raised another $2 billion in April 2025, reportedly reaching a valuation of $30–32 billion despite having no product.

In July 2025, Gross left to join Meta’s AI division. Sutskever then took over as CEO, continuing SSI’s research-first push toward building safe superintelligence.

The post Ilya Sutskever Changes His Mind About How to Build Superintelligence appeared first on Analytics India Magazine.

G20 Johannesburg became a turning point for global tech governance, with India driving outcomes in the US’s absence and shaping the Johannesburg Consensus on open tech and public digital infrastructure.

The Global AI Compact

India used the summit to push for new global rules on AI. Prime Minister Narendra Modi called for a shift away from “finance-centric” to “human-centric” systems, placing human oversight at the core of AI governance.

He proposed safety by design, algorithmic transparency, limits on deepfakes and encouragement for open-source AI models, urging leaders to focus on building the “capabilities of tomorrow” by moving beyond protecting the “jobs of today”.

PM Modi also announced the AI Impact Summit for February 2026, themed ‘Sarvajan Hitay, Sarvajan Sukhay’, which translates to ‘for the welfare of all, for the happiness of all’.

IBSA’s Digital Return

IBSA (India, Brazil, South Africa) re-entered the global tech conversation with the IBSA Digital Innovation Alliance. India offered UPI to Brazil and South Africa, Brazil is exploring linkages with PIX, while South Africa is using the alliance to modernise its payments landscape.

CoWIN will become a shared public health platform, and the bloc will join hands on cybersecurity protocols, women-led technology projects and NSA-level coordination, acknowledging the merger of digital and national security.

The ACITI Partnership

India quietly sealed the Australia-Canada-India Technology and Innovation (ACITI) Partnership, linking the three countries on critical minerals, clean energy, AI, lithium recycling, 6G, quantum and supply-chain resilience.

Canada and Australia bring deep mining strength. India brings its growing AI base. The deal also marks a reset in India-Canada. The pact encourages joint work on recycling lithium, designing cleanties, with operational meetings starting in early 2026.

UAE’s Billion Dollar AI Push for Africa

The United Arab Emirates (UAE) announced a $1 billion ‘AI for Development’ fund to accelerate Africa’s digital adoption, supporting smart classrooms, telemedicine and climate modelling tools. It signals how Gulf economies are using shared AI infrastructure as a path to global inclusion.

DPI Goes Global

Digital Public Infrastructure (DPI) moved to the centre of development discussions. PM Modi proposed an Open Satellite Data Partnership to give developing nations access to satellite imagery and measurements from G20 space agencies—from mixture maps for farmers to storm alerts for coastal communities.

India also launched the G20 Africa Skills Multiplier to train one million digital trainers and a Global Traditional Knowledge Repository to protect medical and ecological heritage from “bio-piracy”.

Critical Minerals Circularity

India pushed a shift from extract-and-export to recycling and urban mining for lithium, cobalt and other battery metals. The Critical Minerals Circularity Initiative aims to reclaim materials from used batteries and electronics, cut pollution and build cleaner domestic supply chains, aligning with Africa’s mineral boom.

Africa’s AI Factories

Strive Masiyiwa, founder of Cassava Technologies, announced five AI factories across Africa—a $720 million investment in local supercomputing powered by NVIDIA processors.

Most of the capacity is already pre-booked by universities, banks and startups signalling a significant leap in data sovereignty and local AI development.

China-South Africa Green Tech Plan

China and South Africa outlined a plan for clean industrialisation, focussed on green mining and adding value on the continent.

The goal is to turn Africa’s mineral resources into manufacturing strength and build innovation-ready green industries.

India’s Bilateral Tech Corridors

India advanced several bilateral tracks. Along with Italy, the country introduced a new initiative to track terrorism financing via crypto, payments and the dark web; progress on space. It covered semiconductors, talent mobility and economic security with Japan.

With South Africa, India reviewed cooperation in AI, public digital systems, minerals and education, and IIT’s African campuses. Telecom networks and 6G frameworks were the focus of the India-UK talks.

Tracking the Drug Terror Economy

PM Modi urged the G20 to target fentanyl networks using data analysis and financial intelligence to trace digital money flows behind synthetic drugs.

A New Digital Order

Johannesburg showed a Global South ready to write its own rules. With the US absent, decisions moved faster—and India anchored a model built on DPI, open technology and human welfare.

The Johannesburg Consensus reflects a world choosing neither Silicon Valley’s market-driven approach nor surveillance-heavy alternatives, but a middle path that keeps technology fair, accessible and centred on human needs as India heads toward the 2026 AI Impact Summit.

The post India Writes the G20 Tech Script in Johannesburg Consensus appeared first on Analytics India Magazine.

For years, the optical interconnect inside the data centre, one of the smallest components in the stack, has been an outsized bottleneck. Every AI model, every hyperscale application, every server cluster ultimately depends on how efficiently data moves across machines. But, the industry standard pluggable transceiver, bulky and power-consuming, has barely evolved at the pace of cloud-scale expansion.

That mismatch is the starting point for Rohin Y and his team at LightSpeed Photonics, who are building what they call the first solderable optical transceiver, a component that is 20 times smaller and five times lower in power consumption than the conventional pluggable version. “We wanted to rethink the interconnect from first principles,” Rohin says. “Not shrink it, reinvent it.”

Their 400 Gbps transceiver, engineered to fit onto a motherboard like any ordinary chip, is being positioned as the “first and only” such product in the global market. Instead of embedding photonics inside the silicon, their architecture keeps it modular—simplifying manufacturing, lowering cost, and dramatically reducing time to market. “We’re a frugal alternative to silicon photonics,” Rohin says. “Others put optics inside the chip; we put it where it’s easier to scale.”

A Market Defined by AI and Power Constraints

The timing could not be more strategic. The world’s data centres, especially those powering AI clusters, are running into severe energy, cooling, and space constraints. Optical interconnects, though critical, are still expensive and cumbersome to integrate. With AI workloads expected to double power consumption every three to four years, the pitch for a smaller, lighter, cooler interconnect is resonating.

The global data centre interconnect market itself is projected to grow into a multi-billion-dollar opportunity, driven by hyperscalers racing to keep up with next-generation AI model demands. But, instead of chasing hyperscalers directly, the company is taking a practical route: OEM partnerships. That means working with players like Supermicro, Quanta, Dell, HP, and Juniper Networks, who already serve the major cloud providers.

“OEMs are the real gatekeepers of data centre hardware,” Rohin says. “If they approve you, the entire market opens.”

The company already has one pilot completed and two more in the pipeline, which will be executed with the fresh infusion of Series A capital.

Business Model

The business model is rooted in a simple, high-volume logic of becoming the go to optical interconnect component for OEMs. The solderable design eliminates custom packaging, reduces touchpoints, and allows the component to integrate directly into server motherboards or network cards.

Co-founder and CTO P V Ramana explained that Vertical-Cavity Surface-Emitting Laser (VCSEL)–based optical interconnects now command over half of data centre deployments because they consume less energy and are cost-effective.

He added that when these components are packaged as compact, solderable units, they can significantly increase data throughput, simplify PCB design, and help lower both power usage and GPU temperatures.”

Mass manufacturing is expected in 2027, and the company is projecting $30 million in revenue that year, with scaling to follow as OEM partnerships deepen. “We’re pre-revenue, but the ramp after OEM qualification is very steep,” Rohin says.

LightSpeed Photonics operates in a global landscape that includes players like Celestial AI, Ayar Labs, and Ranovus, each tackling data bottlenecks in AI and high-performance computing. While many competitors focus on chip-level or co-packaged photonics backed by significant global investment, LightSpeed takes a different route with a solderable optical module that fits into existing system architectures, offering comparable performance through an alternate integration approach.

The Funding Arc

The company has raised $6.5 million in a pre-Series A round, led by pi Ventures. Other investors participating in this round include 500 Global, Indian Accelerator and returning investors – 8X Ventures & Java Capital and Angels from Bay Area. With this, the company’s total funding stands at approximately $8.5 million, including grants.

“LightSpeed’s high speed optical interconnect product family is engineered to address a major hurdle of data movement in large-scale AI and cloud computing, by delivering significantly faster data transfer, and at the same time drastically reducing energy consumption,” said Manish Singhal, founding partner at pi Ventures.

Singhal said, “Their (LightSpeed’s) innovation allows system upgrades without redesigning entire architectures – a game-changer for hyperscale data infrastructure.”

“We’re excited to back a deep tech team that’s solving a global-scale infrastructure challenge from India,” he added.

“Investors saw we weren’t just building a lab prototype,” Rohin says. “We were solving a real data centre problem at the right time.”

The Series A funds have a focused purpose: executing the pilots and establishing a small R&D facility that accelerates iteration cycles. Beyond this, the company is already planning its next fundraise — $20-25 million, supplemented by 50-60% value add through government incentive schemes.

Building in India, Scaling in the US

The company’s go-to-market strategy is two-pronged. US for technology approval because nearly 40% of the world’s data centres are there, and OEM qualification often begins in Silicon Valley. India and developing economies for volume production – especially the Middle East and emerging markets, which are aggressively building new AI-ready data centre capacity.

“We have certain elements that are centred in Singapore for that reason, and we want to gradually bring that capability on a large scale back to India,” says Rohin.

The US is where data centres are and where early traction happens, but the manufacturing and large-scale impact will eventually shift to India, he adds.

Talent Unavailable?

None of this would be possible without the kind of talent the company has managed to assemble. In deep-tech, especially photonics, the global talent pool is small and India’s is even smaller. Yet, the startup has built a team of specialists in optics, RF, electronics, embedded systems, and manufacturing engineering.

The company has also taken a contrarian view on hiring: instead of scaling headcount aggressively. With a 30-member team, they maintain a tight core of engineers capable of working across disciplines. “We look for engineers who can think beyond their specialisation,” Rohin says. “Deep-tech needs intersectional skills.”

The talent function is also tied to India’s growing deep-tech culture. Universities are producing more optics and hardware engineers than ever before. Meanwhile, large tech companies are no longer the only aspirational employers, startups are also building real hardware and are gaining credibility.

“If AI is the engine, the interconnect is the fuel line,” says Rohin. “And right now, the world needs a better fuel line.”

The next year will be defining for LightSpeed Photonics. Closing the Series A, executing the pilots, setting up the R&D centre, beginning the India manufacturing transition, and preparing for the large funding round. All this while pushing toward OEM qualification will test the company’s technical and operational resilience. Rohin says, “there’s no better time to do it.”

The post This Startup is Trying to Fix AI’s Traffic Jam appeared first on Analytics India Magazine.

EON Space Labs said its miniaturised ultra-lightweight space telescope, MIRA, has cleared thermo-vacuum testing, qualifying it for space operations in accordance with NASA standards.

The Hyderabad-based startup confirmed the milestone and said the telescope will fly on a satellite mission in December 2025 along with another Hyderabad space startup, TakeMe2Space.

The test establishes the telescope’s readiness for low Earth orbit and verifies its performance under temperature shifts, vacuum levels, and operational loads. The electro-optical payload weighs 502g and aims to support defence and commercial missions.

Co-founder Sanjay Kumar said, “This is a truly defining moment for us. Space-grade certification proves that ultra-complex, high-precision imaging platforms can now be built and certified entirely from within India.”

This development comes just after the startup raised $1.2 million in a pre-series A funding round in August this year. This was led by MGF Kavachh, with HHV Advanced Technologies joining as an investor and strategic partner.

Along with MIRA, the funding was announced to support the manufacturing of electro-optical and infrared (EO/IR) payloads and the expansion of EON’s engineering team.

The tests took place at an NABL-accredited facility in Ahmedabad. Conditions included vacuum levels below 10⁻⁵ torr and temperature swings from −20°C to +60°C. The telescope also ran imaging and telemetry tasks with an onboard computer, replicating orbital conditions.

“MIRA enables the highest-quality imagery at a fraction of traditional size and weight limitations,” said co-founder Punit Badeka. EON Space Labs is also developing its LUMIRA EO/IR aerial imaging platform for drones, UAVs, eVTOL aircraft and fixed-wing systems.

As the startup aims to build dual-use optical systems for space and aerial missions, co-founder, Manoj Kumar Gaddam, added that the company’s optical and AI-enabled stack aligns with space conditions.

The post Hyderabad’s MIRA Space Telescope Clears NASA-Grade Tests, Set for Launch in December appeared first on Analytics India Magazine.

Cornell Tech has received over $7 million in combined funding from NASA and Schmidt Sciences to overhaul arXiv, the influential open-access research repository that now hosts more than 2.8 million scientific papers.

The investment will accelerate arXiv’s migration to cloud infrastructure, upgrade its ageing codebase and develop new recommendation tools to help researchers better discover relevant preprints.

Greg Morrisett, Jack and Rilla Neafsey, dean and vice provost at Cornell Tech, said the support ensures the platform can scale with the growing needs of the global research community. He called the funding a critical investment in arXiv’s long-term sustainability.

Ramin Zabih, arXiv’s executive director and a professor of computer science at Cornell Tech, noted that the repository is in the middle of a major technology transformation.

The grants, he said, will allow the team to complete the migration without service disruptions, while also experimenting with improved discovery and access tools.

Schmidt Sciences’ contribution will help arXiv expand its engineering capacity and finish the modernisation push.

NASA’s grant will support research into building fairer and more effective search and recommendation capabilities, as well as extending arXiv’s reach into fields aligned with the agency’s priorities, including planetary science.

James Ricci, director of science systems at Schmidt Sciences, said the organisation is eager to help arXiv move to scalable cloud infrastructure and meet rising global demand, calling the platform a vital driver of open science.

Founded in 1991 by physicist Paul Ginsparg, arXiv has grown into a widely used repository for disciplines spanning physics, mathematics, computer science, quantitative biology, finance and more.

Beyond the latest grants, the platform continues to be supported by the Simons Foundation, academic libraries, universities, professional societies and individual donors.

The post Cornell Tech Secures $7 Million From NASA and Schmidt Sciences to Modernise arXiv appeared first on Analytics India Magazine.