‘Every Company Will Have to Become an API Company’

AI agents are quickly becoming the new interface to the internet. While models handle reasoning, APIs let agents act by pulling live data, triggering workflows, and interacting with businesses in real time. As agents move from demos to deployment, APIs are becoming core business infrastructure rather than just developer tools.

For Postman, this shift is familiar territory. Long before AI agents entered the picture, the platform for building and using APIs was built to solve the growing complexity of APIs at scale.

That journey, as co-founder and CTO Ankit Sobti recalled, began not with a grand business plan but with frustration. “I started this as a product in 2012 as a side project,” Sobti said. “It was very much a scratch-your-own-itch problem.”

The growing complexity of working with APIs at scale would eventually turn into one of the world’s most widely used API platforms, now serving over 40 million developers globally.

From a Yahoo Problem to a Global Platform

Postman’s roots lie in Sobti’s experience building APIs inside large organisations like Yahoo.

“We were building an API that every Yahoo vertical depended on—news, sports, finance, the homepage,” he recalled. “We saw the entire lifecycle of building it, operating it, scaling it internally and then exposing it to external customers.”

That experience revealed something deeper than developer convenience. Sobti said they came to understand the challenges of working with APIs not only from a developer’s perspective, but also in terms of how they move the needle for large organisations. It also exposed a gap in the market: the lack of tools focused purely on APIs, not as side infrastructure, but as a first-class product.

Sobti frames APIs as fundamental infrastructure rather than mere technical plumbing.

“APIs are the connective tissue of how the world works today,” he said. “Tens of thousands of developers, across thousands of teams, are building value by using each other’s capabilities.”

This, in his view, is why every organisation is now an API company, whether it realises it or not. “Banks, logistics companies, healthcare, telecom—everyone is opening up APIs,” Sobti quipped. “Either to create new revenue or to support existing revenue.”

The challenge, however, is no longer just building APIs, but managing them at scale.

APIs For AI Agents

Postman is extending its API platform to support AI agent-driven development. The company offers tools to build and test agentic workflows, expose APIs as callable agent tools, and monitor both human and agent usage in real time.

These features include a natural language agent mode, Model Context Protocol (MCP) integration and enterprise observability through Postman Insights.

“Large language models are trained on historical data,” Sobti said. “But APIs allow them to operate in the present moment.”

He offered a simple example: an e-commerce support agent that needs access to real-time shipping status, order details and multilingual responses. According to Sobti, the core challenge is managing the sprawl of APIs across organisations. These APIs must be structured and governed so they can be safely exposed as MCP tools, allowing AI agents to interact with systems reliably and deliver real customer experiences.

He added that this is where Postman is investing heavily—in API catalogues, testing, governance and tools that allow APIs to be exposed reliably to AI-driven workflows.

How AWS Fits Into Postman Strategy

As Postman prepares its platform for agent-driven workflows, partnerships with cloud providers remain central to its strategy. Sobti pointed to the company’s long-standing relationship with Amazon Web Services as an important part of that effort.

“We’ve had a long history in partnership with AWS and are doubling down on that,” Sobti said. Postman is among the early users of AWS’s latest tool, Kiro Powers.

Sobti explained that Kiro Powers works alongside Postman to reduce repetitive setup work in API development and testing. Instead of manually creating requests and configuring environments, Kiro can evaluate an existing Postman workspace and generate a complete API collection with the required endpoints.

“For you to be able to configure Postman, manage tests and manage workspaces from within Kiro Powers itself is fascinating,” Sobti said.

Looking Ahead

​​One of Postman’s biggest advantages, according to Sobti, is its user base.

“Many users inside organisations who are using Postman, who love the product, who are trained on using the product and using APIs as well. So Postman becomes a very effective distribution channel.”

That developer-first adoption has helped Postman evolve from a tool into a platform used by individuals, teams and enterprises alike.

“We’re at the precipice of fundamentally new consumer experiences,” Sobti said. “We don’t yet know the winning form factor, but conversational agents are clearly one direction.”

What he is certain about is the role APIs will play.

“If agents are how users interact with businesses, and you don’t have APIs to support that, it’s going to be very hard,” Sobti said. “Every company will have to become an API company.”

And in that future, agents may well become the fastest-growing API consumers of all.

The post ‘Every Company Will Have to Become an API Company’ appeared first on Analytics India Magazine.

Why Scrapping 10-min Delivery Doesn’t Give Gig Workers Much Relief

India’s gig economy has fundamentally changed consumer behaviour, but its benefits may only be limited to app users and the companies that built them. Millions of delivery partners, who are the backbone of quick commerce, are squeezed in the middle amid relentless pressure, penalties, and abysmal pay.

According to NITI Aayog, India had 77 lakh e-commerce and quick commerce delivery workers in 2020-21, and this number is projected to soar to 2.35 crore by 2030. Swiggy reports about 5.4 lakh delivery partners across food delivery and Instamart, while the Zomato-Blinkit ecosystem together engages roughly 7-8 lakh monthly active delivery partners and up to ~15 lakh unique riders annually due to high churn.

The situation has boiled over. On December 25 and December 31 last year, gig-worker unions declared a nationwide strike, demanding scrubbing 10-minute delivery and reinstating previous payout structures. They contended that such demanding timelines compromise safety that has resulted in accidents and fatalities.

Mohammed Ali (name changed), a delivery partner for Zepto, tells AIM that the platforms have strict targets that the riders need to meet to earn incentives. “They say, if you complete 30 orders, you will get a ₹500 incentive. That’s why we end up rushing. The incentive pushes us to complete as many deliveries as possible in a day, which in turn pushes the 10-minute delivery mindset.”

The government and parliamentarians have been stirred to action.

Raghav Chadha, a member of Parliament from the Aam Aadmi Party, stepped into the shoes of a Blinkit delivery rider for a day, highlighting the everyday realities of gig workers in a viral video. A delivery partner talked to him about getting fractures on the job, paying penalties, and earning very little from the platform.

However, Deepinder Goyal, founder of Zomato and Blinkit, recently claimed that delivery partners are under no pressure. “Our 10-minute delivery promise is enabled by the density of stores around your homes. It’s not enabled by asking delivery partners to drive fast. Delivery partners don’t even have a timer on their app to indicate what was the original time promised to the customer,” he posted on X.

Much Needed Relief

After a meeting on January 13 with the union minister of labour and employment, Mansukh Mandaviya, Blinkit scrapped its 10-minute delivery claim. Following suit, Zepto and Swiggy’s Instamart also announced they would remove their 10-minute delivery timelines. The companies also agreed to remove the claims from their branding and advertising.

Satyamev Jayate. Together, we have won..
I am deeply grateful to the Central Government for its timely, decisive and compassionate intervention in enforcing the removal of the “10-minute delivery” branding from quick-commerce platforms. This is a much needed step because when…

— Raghav Chadha (@raghav_chadha) January 13, 2026

The Gig Workers Association (GigWA) has welcomed the rollback of the 10-minute delivery guarantee, acknowledging the dangerous pressure that such intense timelines place on delivery workers.

In a statement, the association noted that the 10-minute delivery model forced workers to rush, put themselves at risk on the roads, and work long hours under constant app-driven pressures from rewards, ratings, and task assignments.

In conversation with AIM, a delivery partner for Swiggy Instamart, though delighted that the 10-minute mandate has been scrapped, still wants the Supreme Court to intervene.

Has Anything Changed on the Ground?

However, scrapping the 10-minute delivery structure may not change the delivery partners’ reality overnight. An X user posed a question, “Quick-commerce companies like Blinkit, Zepto, and Instamart promise 10-minute deliveries. But how much of a city can they actually reach in that time?”

He mapped the quick commerce stores in Bengaluru, Mumbai, Delhi, Hyderabad, and Pune, highlighting real routing data and actual coverage zones across major urban hubs for Blinkit, Zepto, and Instamart.

The secret to 10-minute delivery isn't driver speed. The average delivery distance is shorter than your morning walk. Dark stores are everywhere, hiding in plain sight.
I mapped quick-commerce infra in Indian cities. Every dark store, real routing data, actual coverage zones… pic.twitter.com/2Ob0bDwi4u

— Anup (@anupbhat30) January 2, 2026

AIM cross-verified how long it would take for an order to be delivered. While Blinkit still claims a 10-minute delivery, Swiggy/Instamart claims delivery in 5 minutes. Blinkit and Swiggy have also not responded to AIM’s request for comments.

Similar concerns were echoed by several X users.

“[I work mostly with] Zepto. I used to work with Zomato a lot, but I stopped because we had to travel very long distances. For Zomato, sometimes we have to go 4 km for just ₹30. We go 4 km, then come back another 4 km, all for ₹30. Even for snacks, if the total distance is 4 km, they give only ₹60 at the end,” Ali explains.

He says that with Zomato, the distances are long, but the pay is low. In contrast, with Zepto, the distances are shorter, allowing him to earn more. He explains that if he works 12–13 hours for Zepto, he can earn ₹1,500; however, he won’t make that much with Zomato even if he works for 14 hours.

Last year, Amazon too jumped on the 10-minute delivery bandwagon with Amazon Now, opening more than 100 micro-fulfilment centres across Bengaluru, Delhi, and Mumbai. However, the company declined to comment on scrapping 10-minute deliveries.

Nonetheless, it remains unclear how scrapping the 10-minute delivery guarantee will benefit riders, especially since the dark stores are situated very close to residential areas. As a result, delivery partners often earn less because their pay is based on distance travelled, and there is little incentive to complete the maximum number of deliveries in a day.

Ali notes that wages were higher earlier. “I don’t know why it happened automatically. Earlier, per order, the app would show ₹42, ₹36, or ₹30 as power pay. Now it shows ₹30, but when we actually get the money, it’s only ₹16 or sometimes just ₹6.”

Issues also persist with accident insurance. Not wishing to be named, the Swiggy Instamart delivery partner quoted above claims the platform does not provide any financial support for safety gear, such as helmets, nor does it reimburse fuel expenses. He says colleagues who suffered bone fractures in accidents while delivering orders had to incur medical costs ranging from ₹1.2–1.8 lakh, borrowing money from their families.

“For us, coming from Bihar to a big city like Bengaluru, even minor treatments cost us a lot of money, especially when we are also bearing all the costs of transit, fuel, and safety gear,” he laments.

He says that although Swiggy earlier stated that it would cover his treatment costs for minor injuries, ultimately, they did not pay him anything. However, Swiggy asserts that it offers insurance benefits.

“All our delivery partners on Swiggy get insurance benefits like accidental coverage of ₹2 Lakh, accidental death and disability cover of ₹10 Lakh, accidental OPD of ₹10,000, loss of pay compensation up to 3 months in case of accident, and free, on-demand ambulance service right from their first order,” the company wrote in its blog.

In a report titled ‘Partners in Peril: Testimonies From Platform Workers’, GigWA mentioned India’s economy will be best served if a proper legal framework is put in place to regulate the online economy and improve the working conditions of gig workers.

The post Why Scrapping 10-min Delivery Doesn’t Give Gig Workers Much Relief appeared first on Analytics India Magazine.

Inside OpenAI’s $10 Bn Shortcut to Real-Time AI

As demand for real-time AI applications grows, the focus is turning to inference infrastructure. Low-latency performance is emerging as a key bottleneck in building applications like coding agents and voice-based interactions, forcing AI developers to look beyond traditional GPU-heavy architectures.

Real-time inference is critical for AI models to make instantaneous decisions, fuelling real-time applications such as autonomous driving and financial fraud detection.

OpenAI now has a first-mover advantage after entering a multi-year partnership with AI chipmaker Cerebras to deploy 750 megawatts of wafer-scale AI systems for inference. The rollout will begin in 2026 in multiple phases, with the infrastructure designed to serve OpenAI customers globally. The deal is valued at more than $10 billion, according to the Wall Street Journal.

“Cerebras adds a dedicated low-latency inference solution to our platform. That means faster responses, more natural interactions, and a stronger foundation to scale real-time AI to many more people,” Sachin Katti of OpenAI had said in a statement.

The partnership builds on years of engagement between OpenAI and Cerebras, with Sam Altman one of Cerebras’ early investors. The deal comes at a critical moment for OpenAI as it diversifies its AI infrastructure. It also comes on the heels of Apple and Google’s surprise partnership to infuse Google’s AI tech into iOS, including in the updated version of Siri.

Pressure on NVIDIA

Competition in the AI inference market is intensifying as AMD and Intel build lower-cost alternatives to GPUs and hyperscalers like Google and Amazon build their own TPUs.

In this heated environment, OpenAI seems to have struck gold.

Cerebras claims its systems can run large language models up to 15x faster than GPU-based alternatives. The company’s most recent and current chip architecture is the Wafer Scale Engine-3 (WSE-3), which powers its latest systems such as the CS-3, a wafer-scale AI processor with around 4 trillion transistors and roughly 900,000 AI-optimised cores.

According to Cerebras, the CS-3 system is up to 21x faster than NVIDIA’s DGX B200 Blackwell GPU and operates at about one-third the cost and power, supporting applications including conversational AI, real-time code generation and reasoning tasks.

In an exclusive conversation with AIM in October last year, Andrew Feldman, co-founder and CEO of Cerebras, said wafer-scale computing sits at the heart of the company’s next phase of growth. “This is the largest chip in the history of the computer industry,” he boasted, adding that by keeping far more data on a single chip, Cerebras can process information faster, move data less frequently, consume less power, and deliver results in far less time.

“AI becomes exciting when the response is real time,” Feldman noted. “Nobody wants to wait 40 seconds or four minutes for an answer.”

NVIDIA is not a bystander either. Recently, Groq, a US-based company that builds specialised hardware for AI inference, announced a non-exclusive licensing agreement with NVIDIA valued at about $20 billion. As part of the deal, Groq founder Jonathan Ross, president Sunny Madra, and several other employees joined the company, bringing with them Groq’s low-latency language processing unit processor.

Economics of Inference

Beyond speed, the economics of inference are central to Big Tech’s AI strategy. Faster inference can translate into lower cost per token by reducing compute time, energy consumption, and infrastructure overhead.

“B200-class GPUs can be cost-effective when utilisation is high, traffic can be deeply batched and the software stack is well optimised,” Carmen Li, CEO of Silicon Data, tells AIM.

Li adds that many interactive inference workloads such as chat, agents and voice are bursty and sensitive to latency, which limits batching and creates inefficiencies. “These workloads don’t behave well on heavily batched systems,” she says.

Batching involves grouping multiple data inputs to process them together as a single batch to boost computational throughput.

Li notes that wafer-scale systems perform better economically in such scenarios by reducing the need for multi-GPU coordination and interconnect overhead, consolidating compute and memory bandwidth into a single system, and delivering more predictable latency when strict service-level requirements must be met.

Feldman highlights that GPUs still make sense for slower, throughput-oriented tasks like synthetic data generation. But for agentic AI, real-time reasoning, and customer-facing applications, wafer-scale has a decisive edge.

Escaping CUDA Lock-in

One of the biggest barriers to moving away from GPUs is software lock-in, particularly around NVIDIA’s parallel computing platform CUDA. Cerebras claims it has largely eliminated that friction.

“The way you move quickly and disintermediate CUDA is through the use of an API,” Feldman says. “Most application developers don’t want anything to do with CUDA.”

Instead, developers connect to Cerebras much like they would to any cloud AI API, by changing just a few lines of code.

However, Li points to software constraints, saying wafer-scale platforms rely on specialised programming models, compilers and APIs that are narrowly optimised for machine learning. This limits flexibility compared to the CUDA ecosystem and suggests wafer-scale will function as a specialised inference tier rather than a universal replacement for GPUs.

Li notes wafer-scale inference can be faster and more power-efficient for workloads that fit its architecture, but fabrication yield remains a key variable. Even with fault tolerance, wafer-scale manufacturing is difficult, and if yields drive up costs, the performance benefits may not fully offset higher capital expenditure.

She adds that wafer-scale systems do not eliminate the costs of distributed computing. “Once workloads exceed a single system, or when geo-distribution and high availability matter, familiar scaling penalties reappear,” Li notes, adding that the approach primarily optimises single-node latency and efficiency rather than large-scale distributed inference.

What’s Next for Cerebras

Cerebras is reportedly in talks to raise $1 billion at a valuation of about $22 billion, nearly tripling its previous valuation. Last September, the company raised $1.1 billion in an oversubscribed Series G funding round, valuing it at $8.1 billion post-money.

In addition to OpenAI, Cerebras works with Abu Dhabi-based AI group G42. The company filed confidentially for an IPO in September 2024 but withdrew the filing in October 2025 amid scrutiny from the Committee on Foreign Investment in the United States over its ties to G42.

Other than OpenAI, Cerebras’ customers include AWS, Meta, IBM, Mistral, Cognition, and Hugging Face.

The post Inside OpenAI’s $10 Bn Shortcut to Real-Time AI appeared first on Analytics India Magazine.

Google Launches TranslateGemma, Takes On ChatGPT Translate

Google on January 15, announced TranslateGemma, a new suite of open translation models built on Gemma 3, to support text translation across 55 languages.

The models are available in 4B, 12B, and 27B parameter sizes and are intended for use across mobile, local, and cloud environments, the company said. According to Google, TranslateGemma offers higher translation quality with fewer parameters by distilling capabilities from its larger Gemini models.

“By distilling the knowledge of our most advanced large models into compact, high-performance open models, we have created a suite where efficiency doesn’t require a compromise on quality,” the company said in its announcement.

TranslateGemma models are available through multiple channels, including Kaggle, Hugging Face, Vertex AI, and Google’s Gemma Cookbook.

The models were trained and evaluated across 55 language pairs, covering high-, mid-, and low-resource languages. Google said TranslateGemma reduced translation error rates across all tested languages compared to the baseline Gemma models. In addition, the company trained the system on nearly 500 more language pairs to allow researchers to fine-tune models for specific use cases.

Google said internal tests showed the 12B TranslateGemma model outperformed the larger Gemma 3 27B baseline on the WMT24++ benchmark using the MetricX framework. The 4B model delivered performance comparable to the 12B baseline, making it suitable for on-device inference.

Meanwhile, OpenAI has also introduced ChatGPT Translate, a web-based translation tool. The service is positioned as an alternative to Google Translate and uses ChatGPT models to translate text across more than 50 languages. It focuses on text input with automatic language detection and can be used without an account, though signing in enables additional features.

The tool uses a dual-panel layout, with one section for input text and another for the translated output. Users can select source and target languages from dropdown menus. ChatGPT Translate also allows users to adjust the tone of translations, such as making them more formal, simplified, or suited for academic use.

The post Google Launches TranslateGemma, Takes On ChatGPT Translate appeared first on Analytics India Magazine.

RCB to Collaborate with Staqu for AI-Enabled Crowd Management at Chinnaswamy Stadium

Royal Challengers Bengaluru (RCB) has proposed the deployment of advanced AI-enabled surveillance systems at the M. Chinnaswamy Stadium in collaboration with video analytics firm Staqu.
RCB has committed to bearing the entire one-time cost of the proposed initiative, estimated at approximately ₹4.5 crore. The deployment is intended to enhance overall safety standards while ensuring a smoother and more secure matchday experience for fans.

In a formal communication to the Karnataka State Cricket Association (KSCA), RCB outlined plans to install between 300 and 350 AI-powered cameras across the stadium premises. The system will leverage real-time video analytics to monitor crowd movement, regulate queueing, detect unauthorised access, and track entry and exit points to enable quicker response by law enforcement agencies.

The AI platform, developed by Staqu, integrates video, audio and text analytics to support faster and more accurate incident detection. Its real-time capabilities allow early identification of situations such as intrusions, access violations and potential security threats, helping authorities intervene proactively.

Staqu has previously worked with multiple State Police departments, providing facial recognition and intelligent monitoring solutions for crowds, perimeters, vehicles and objects as part of routine surveillance and investigation processes.

Staqu’s flagship product is JARVIS, an AI platform that analyses video and audio feeds to support security and crowd monitoring. Instead of offering separate tools, JARVIS combines features such as crowd analysis, intrusion detection, facial recognition and audio monitoring into a single system.

Jarvis GPT, launched in February 2025, works as an added AI layer that allows authorised users to access and analyse surveillance data using natural language commands for the retail industry.

The post RCB to Collaborate with Staqu for AI-Enabled Crowd Management at Chinnaswamy Stadium appeared first on Analytics India Magazine.

Why the PSLV Setback Didn’t Ground India’s Space Startups

On the morning of January 12, ISRO’s PSLV-C62 lifted off from Sriharikota carrying EOS-N1, a hyperspectral earth observation satellite, along with a cluster of 15 other payloads. The mission reflected India’s growing space ecosystem, with satellites from private Indian players and a handful of international ones.

Three Indian startups were eager to demonstrate a proof-of-concept for space technologies that had taken years to develop. The launch was meant to hit a milestone.

But soon after the take-off, the mission fell short.

Indian Space Research Organisation (ISRO) detected a third-stage anomaly, cutting the flight short before the payloads could be deployed. Satellites meant to activate in orbit never separated from the rocket. They got lost from the intended trajectory.

For founders watching this telemetry, a new, stark reality came to the fore. The mission ended early.

The aftermath was quieter, faster, and more revealing than the launch. While the loss of hope was immediate, the recovery began just as fast.

ISRO chairperson V Narayanan reported that the vehicle encountered disturbance roll rates and deviated from its flight path. “We are analysing the data, and we shall come back at the earliest,” he said.

Post this, engineers began assessing what failed. Founders had to call customers and investors to recalibrate timelines. Regulators issued statements of support. Not many spoke of the panic and roadblocks that these startups might face.

The impact went beyond startups, prompting ISRO to reassess future missions. What does recovery look like, and what does this reveal about India’s private space sector?

“It was rather unfortunate,” Bhaskar Majumdar, managing partner at Unicorn India Ventures (UIV), told AIM. “The PSLV is the workhorse of the Indian satellite programme and has one of the lowest failure rates of 6%.”

Majumdar, whose firm has invested in deep tech and space tech startups, noted that the timing was unsettling for many in the ecosystem. To date, the PSLV rocket has been used in 60 launches, with two aborted missions.

Last year, during the launch of the EOS-9 surveillance satellite into the intended orbit, its launch vehicle, PSLV-C61, also encountered a similar technical issue.

It is ironic that both missions (C-61 and C-62) failed within eight months and faced similar problems. Majumdar pointed out that the timing was unsettling for many in the ecosystem.

“It is a matter of concern for ISRO as this not only slows down startups but becomes a blot in the space programme, especially when we are doing complex missions like Mangalyaan and Chandrayaan,” he added.

India’s space startup ecosystem is no longer experimental. It is commercial, customer-facing, and capital-intensive. A launch failure of this sort not only delays science but also pauses revenue, pilot projects, and contracts. And Indian space startups are no longer optimising only for the success of their satellites or payloads but for recovery in such scenarios.

What Breaks & What Survives

TakeMe2Space (TM2S), a Hyderabad-based space startup, sent its 14 kg MOI-1 satellite on this launch. MOI-1 aimed to be an ‘AI lab in space’ for about 15 customers. This satellite also featured EON Space Labs’ miniaturised, ultra-lightweight space telescope, MIRA, which had recently passed thermo-vacuum testing, qualifying it for space operations in accordance with NASA standards.

Since no launch sequence was initiated, MOI-1 was not ejected and therefore did not power on,” TM2S founder and CEO Ronak Kumar Samantrayv said.

MOI-1 was designed to test an ambitious goal. The satellite aimed to run AI models directly in orbit, reducing dependence on ground processing. Samantray said four customers were prepared to push the AI models into satellite mode. He added that this process is now on hold until they can get the satellite into orbit as quickly as possible.

What did not pause was rebuilding.

The startup had already prepared for failure scenarios during assembly. The team built multiple versions of every subsystem, partly to manage late-stage risks. That decision has now shaped the recovery timeline.

Samantray stated that they have a complete set of satellite components on hand and expect to assemble a new MOI-1 by mid-March. They are also considering alternative launch options, such as ISRO’s SSLV rocket, with plans to return to orbit before August.

The setback still carries a cost. Customer pilots are delayed, and the demonstration data for TM2S has not been generated. This issue is common among startups that sent payloads, including Hyderabad-based Dhruva Space, which supported five payloads on the mission covering satellites, launch coordination, and ground systems.

Sanjay Nekkanti, CEO and co-founder of Dhruva, stated that their current goal is a prompt and steady turnaround, enabling both their team and clients to prepare for upcoming launches in less than a few weeks.

Dhruva’s model heavily emphasises the ground segment. While satellites might fail to reach orbit, ground infrastructure remains operational. That distinction is important when customers are already managing systems.

According to investor Padmaja Ruparel, co-founder of IAN Group, most Dhruva customers on the mission already had ground infrastructure delivered and operational. That infrastructure will support future launches without requiring replacement.

From a programme perspective, this reduces the cost of failure. Ruparel added that Dhruva is already working with its customers to announce the next set of missions soon.

Bengaluru-based OrbitAID was another Indian startup that launched its AyulSAT payload onboard this mission. It was India’s first tech demonstrator for in-orbit satellite refuelling, testing the transfer of propane fuel in space to extend satellite life, reduce debris, and build an ‘on-orbit economy’ using its unique docking interface.

This 25kg payload aimed to validate fuel, power, and data transfer, with future plans for a chaser satellite for full ship-to-ship docking. Had this PSLV mission succeeded, AyulSAT could have marked a milestone for India in on-orbit servicing and space sustainability.

Patience, Capital, and Confidence

For investors backing space startups, launch anomalies test more than technology. They test conviction. Majumdar characterised deep tech investing as demanding a different mindset. He noted that failures are not re-evaluated after they occur but are considered a normal part of the growth process.

He noted that investor response is crucial during failure, as entrepreneurs face pressure from customers, teams, and cash flow, and that a withdrawal then worsens the damage.

UIV plans to continue backing affected companies, including TakeMe2Space and OrbitAID, if short-term cash flow pressure emerges.

Majumdar stated that increasing the number of launch sites and operations is necessary to reduce dependence on ISRO. He highlighted the need for an accelerated launch schedule in the near future, given the large number of space startups and the VCs interested in this domain.”

Private players such as Agnikul Cosmos, Skyroot Aerospace and EtherealX will help, but the transition will take time.

Ruparel echoed a similar sentiment. “We are helping our companies with customer traction and retention through technology adoption, scale, risk & execution speed,” she said, adding that deep tech is a bet for the long run.

She viewed failures as part of industrialisation, not breakdown. “In spacetech, thousands of satellites will be built over the next decade,” Ruparel said. She argued that on this scale, failures are inputs to speed rather than bottlenecks.

Earlier, Indian space startups operated under a scarcity model, with rare launches and long waits after failures. Today, startups plan relaunches within months. They design satellites for quicker rebuilds, begin insurance talks earlier, and typically include contingency clauses in customer contracts.

Another fallout from the PSLV-C62 anomaly has been the spotlight on a gap that many Indian space startups now regret: none of their satellites on this mission were insured. Recent reports show that these Indian companies didn’t obtain launch or satellite insurance mainly because premiums are high, and affordable, tailored space policies aren’t available in India.

This gap contrasts with global practice, where launch insurance is standard for most commercial payloads. Analysts and insiders have pointed out that the lack of an accessible insurance market in India leaves startups exposed to total losses when missions fail, especially for first-of-a-kind or experimental satellites.

Will the Government Step in?

For regulators and institutions, the PSLV anomaly arrived at a sensitive moment, as India is pushing to open its space sector while preserving confidence in its national launch capabilities.

According to the year-end review for 2025 provided by the Department of Space in December, India’s space programme was already riding a wave of milestones before the PSLV anomaly.

Last year, ISRO celebrated its 100th launch from the Sriharikota spaceport, marking a milestone that represents nearly fifty years of operational growth. It also revealed plans to expand its launch infrastructure with a third launch pad at Sriharikota to support next-generation rockets and increased launch frequency.

IN-SPACe, the government body tasked with enabling private space activity, responded by emphasising support rather than caution. Dhananjay Khot, director of strategy and planning at IN-SPACe, described an organisation-wide effort to support non-government entities working through recovery.

“I have no doubt that this setback will be turned into another triumph for the Indian space sector. This is merely a delay, and not a denial of their dreams and plans,” he said.

In the interim, ISRO’s schedule is critical as more startups prepare for flight, lengthening launch queues. This tension is common in mature space economies. Last year, NASA’s Lunar Trailblazer, a small lunar orbiter, lost contact and power shortly after its February 2025 launch due to spinning that prevented solar panel orientation.

The organisation ended its lunar water-mapping mission in July 2025. In April 2025, a Firefly Aerospace Alpha rocket launch failed, losing its payload, and in May 2025, a SpaceX Starship test flight experienced an onboard leak and loss of control.

For now, Indian space startups aren’t waiting for perfect conditions, but building resilient systems that can endure setbacks. Despite PSLV C-62’s failure to complete its mission, the ecosystem surrounding it continues unabated.

The post Why the PSLV Setback Didn’t Ground India’s Space Startups appeared first on Analytics India Magazine.

AI Video Startup Higgsfield Raises $130 Mn in Series A, Reports $200 Mn ARR

AI video startup Higgsfield announced on January 16 that it has secured an $80 million Series A extension, with investments from Accel, AI Capital Partners (the US arm of Alpha Intelligence Capital), Menlo Ventures, among others, bringing the total Series A funding to over $130 million and valuing the company at more than $1.3 billion.

Higgsfield CEO Alex Mashrabov announced that the new funding will support the global expansion of AI models for advertising, marketing content, and music videos, as well as ongoing R&D. The company also aims to improve its API and marketing automation for clients, creating high-capacity marketing content systems.

This funding round comes after Higgsfield achieved a $200 million annual run rate in less than nine months, doubling from $100 million in just about two months.

Since its launch in April 2025, the platform has gained over 15 million users globally and currently generates 4.5 million videos per day. Higgsfield is transforming marketing production through high-quality, automated creative generation at scale, resulting in over three billion social media impressions, making it one of the most popular generative AI platforms in terms of social media presence.

“Traditional video production wasn’t built for the pace modern marketing demands,” Mashrabov said in a statement. “In that world, a 16-year-old with taste can outperform a studio pipeline, because on social media the advantage goes to what earns attention and converts, not what took the longest to produce.”

Higgsfield reports that 85% of its users are social media marketers, with 80% already producing commercial work, indicating that platform adoption has matured beyond casual content creation. A key insight is the rapid uptake of generative video among marketers, who are now managing full workflows, from ideation to publishing, within a single system.

Moreover, many direct-to-consumer advertisers are transitioning to a generative AI-first model, using automation pipelines like URL-to-ad to quickly create on-brand video variations from product pages. Some customers using Higgsfield’s beta marketing automation product are reportedly investing over $200,000 annually.

Jeff Herbst, a Higgsfield board member and former head of corporate development at NVIDIA, said the company’s adoption signals a move from pilots to embedded production use.

“When a platform moves beyond pilots and into daily production across enterprises, the outcome is clear,” Herbst said.

The post AI Video Startup Higgsfield Raises $130 Mn in Series A, Reports $200 Mn ARR appeared first on Analytics India Magazine.

Wipro Q3 Net Profit Falls 7% YoY Despite Margin Expansion

WiproWipro

For the quarter ended December 31, 2025, Wipro Limited reported steady revenue growth. However, profit declined due to one-time labour related charges.

The IT giant posted revenue of ₹23,555 crore in Q3. This marked a 5.5% year on year increase from ₹22,318 crore in the same quarter last year. Revenue also rose sequentially by about 4% from ₹22,697 crore reported in Q2FY26. It indicates a modest recovery in business momentum.

Net profit for the quarter fell to ₹3,119 crore from ₹3,358 crore a year earlier, a 7.11% decline. It fell short of analyst expectations. It reported a one-time charge of ₹300 crore linked to the implementation of India’s new labour codes.

The quarter highlights Wipro’s ability to deliver incremental revenue growth and outperform market sales expectations, even as profitability was weighed down by regulatory costs and a challenging demand environment.

For Q3, Wipro’s peers, including TCS, Infosys and HCLTech, also reported a profit decline because of ongoing labour code impacts.

CEO and managing director Srini Pallia at Wipro said the company witnessed an increase in the adoption of its AI-led offerings. He said Wipro Intelligence is emerging as a differentiator and contributed to several deal wins during the quarter. The tech giant scaled AI-led delivery through internal platforms, including WINGS and WEGA.

Wipro expanded its global innovation network to support AI-driven execution.

CFO Aparna Iyer said the margin expansion reflected sustained focus on execution rigour. She highlighted the strong cash generation during the quarter and said the board declared an interim dividend of ₹6 per share, taking the total payout for the year to $1.3 billion.

The company’s voluntary employee attrition stood at 14.2% on a trailing 12-month basis. It remains stable compared to the higher churn levels seen over the last few quarters.

Looking ahead, Wipro offered a cautious outlook for the March quarter. The company’s growth guidance was pegged at 0% to 2.0% on a constant-currency basis. The guidance reflects ongoing uncertainty in client spending, even as interest in AI-led transformation continues to grow.

The post Wipro Q3 Net Profit Falls 7% YoY Despite Margin Expansion appeared first on Analytics India Magazine.

Wipro Adds 6,529 Employees in Q3 as Attrition Eases

Wipro increased its workforce in the December quarter. It added 6,529 employees in Q3 FY26, as fresher onboarding picked up pace amid modest improvement in business activity. The company’s total headcount stood at 2,42,021 at the end of the quarter.

The rise in employee numbers comes even as the company reported a decline in profitability during the quarter. Net profit fell 4% quarter-on-quarter and 7% year-on-year to ₹3,119 crore. Its IT services operating margin improved to 17.6%.

Attrition moderated slightly during the quarter, easing to 14.2% on a trailing 12 month basis from 14.9% in the previous quarter. It indicates a gradual stabilisation in employee churn after prolonged pressure across the sector.

Earlier, the management had indicated that hiring plans for FY26 would remain cautious. While Wipro intends to hire 10,000 freshers through campus recruitment, a number similar to its FY25 target, it said final decisions would depend on how the business environment evolves amid continuing macroeconomic uncertainty.

Operationally, utilisation showed some softening. Wipro’s utilisation rate, excluding trainees, declined to 83.1% in Q3 from 86.4% in Q2. It reflects higher intake of freshers and near-term bench build-up.

The headcount increase stands in contrast to its peers in the industry. TCS reported a decline of over 11,000 employees in Q3, while Infosys added 5,043 employees. It underscores divergent workforce strategies among large Indian IT firms as demand remains uneven.

The company has set up 50 Centres of Excellence across universities to upskill graduates. “We work closely with universities on specific curriculum areas such as AI, cybersecurity, data, and engineering, and hire talent from these programmes,” Saurabh Govil, chief human resources officer, said. “The future will involve an AI-plus-human model, with new roles emerging due to AI.”

That said, the campus recruitment remains muted. The company said it plans to ramp up campus hiring to about 2,500 people in Q4.

Meanwhile, CFO Aparna C. Iyer said margin expansion reflected tighter execution and added that strong operating cash flow, at 135% of net income, supported the company’s financial position during the quarter.

The post Wipro Adds 6,529 Employees in Q3 as Attrition Eases appeared first on Analytics India Magazine.