Shunya Labs Unveils AI Model for India’s Code-Mixed Speech

India AI Voice AgentIndia AI Voice Agent

Shunya Labs has announced the launch of Zero Codeswitch, a speech recognition foundation model designed to understand naturally code-mixed and multilingual Indian speech, a key limitation in existing voice-based AI systems.

The Gurugram-based voice AI company said the model is built to recognise how people in India actually speak, frequently blending Hindi, English and regional languages within the same sentence without relying on intermediate translation layers. The company says that Zero Codeswitch has achieved a 3.10% Word Error Rate (WER) on the OpenASR leaderboard, representing a 48% improvement over the next-best competing model, according to the company.

Shunya Labs said the model is designed to run efficiently on standard CPUs, reducing deployment costs by up to 20 times while maintaining sub-100 millisecond latency for real-time applications such as customer support, voice assistants and automated call centres.

“Shunya Labs was built with a focus on deep research rather than short-term marketing narratives,” Ritu Mehrotra, CEO and co-founder of Shunya Labs said. “With Zero Codeswitch, we are building foundational technology for Indian languages that prioritises accuracy, latency and real-world usability. Our goal is not just to adopt AI, but to build it at the foundation level in India.”

Unlike global speech models that are primarily trained on English data and later adapted for Indian languages, Shunya Labs said its foundation models are trained from the ground up on millions of hours of real-world Indian speech data. This includes variations in accent, dialect, pronunciation and slang across regions, allowing the system to better handle Hinglish and other code-mixed speech patterns.

“‘Shunya’ represents our philosophy of starting from first principles,” Sourav Bandyopadhyay, CTO and co-founder of Shunya Labs said. “With Zero Codeswitch, we are creating an intelligence layer that truly listens, engineered for India’s linguistic diversity and optimised for real-world deployment,.”

The company said Zero Codeswitch is intended for enterprise and public-sector use cases where data privacy is critical. The model can be deployed on-premises or in air-gapped environments, allowing organisations to train domain-specific versions while retaining control over sensitive data. Shunya Labs said it complies with HIPAA, SOC 2 Type II and ISO 27001 standards and supports CPU-first deployments to reduce reliance on specialised GPU infrastructure.

The post Shunya Labs Unveils AI Model for India’s Code-Mixed Speech appeared first on Analytics India Magazine.

AXISCADES Subsidiary Mistral Wins ₹100 Cr Defence Radar Subsystem Order

AXISCADES Technologies’ subsidiary Mistral Solutions has secured a defence contract worth around ₹100 crore for radar subsystems. The order comes from the LLTR Ashwini programme, India’s indigenous low-level transportable radar developed by DRDO and BEL, the company said in a statement.

The contract involves the supply of signal and data processing units and associated spares over a two-year period. The systems will support low-altitude surveillance against UAVs, helicopters and fighter aircraft under the Ashwini radar programme.

Mistral will deliver liquid-cooled processing units designed for radar operations under demanding field conditions. The company said the order provides revenue visibility over the delivery period and strengthens its position in defence electronics programmes led by DRDO and BEL.

Commenting on the order, C Manikandan, CEO of Mistral, said, “LLTR Ashwini programme win reflects Mistral’s continued commitment to supporting India’s defence indigenisation initiatives and contributing critical technologies to next-generation military systems.”

AXISCADES said the win builds on its long-standing work with defence forces, laboratories and public sector units across radar, avionics and electronic warfare domains. The company operates across India, Europe and North America, with a workforce of over 3,000 professionals.

In the defence sector, IG Defence (formerly IG Drones) also recently secured orders from the Indian Army and the Indian Navy for its indigenous counter-drone system, the IG T-Shul Pulse Anti-Drone Gun.

The system will be deployed on the front lines to strengthen India’s counter-unmanned aerial systems capability. This followed the company’s contract with the Indian Air Force signed in March last year.

The post AXISCADES Subsidiary Mistral Wins ₹100 Cr Defence Radar Subsystem Order appeared first on Analytics India Magazine.

Employee Union Flags Fresher Onboarding Delays at Wipro, Seeks Labour Ministry Intervention

The Nascent Information Technology Employees Senate (NITES) has approached the labour and employment ministry, alleging prolonged onboarding delays and misleading recruitment practices by Wipro Limited that have left hundreds of selected graduates in prolonged uncertainty and financial distress.
In a representation dated January 19, addressed to union labour and employment minister Mansukh Mandaviya, NITES said the complaint relates to fresh graduates selected through Wipro’s campus and off-campus hiring, including the Turbo and NextGen Talent programmes.
According to the IT employees’ body, letters of intent (LoIs) were issued around May 2025, confirming selection, roles, compensation structure and onboarding processes.

“In several cases, the company also issued formal onboarding communications confirming a joining date, work location and completion of documentation formalities,” the letter said.
However, NITES alleged that “despite these assurances, the candidates were not onboarded on the promised dates”, even after background verification was completed and company systems showed the status as “100% completed with no action required.”

The union added that affected candidates repeatedly contacted the company through official channels over several months but received either no response or “vague and automated responses referring to business demand, future batches or tentative timelines that were never honoured.”
NITES highlighted the impact on young graduates, noting that many were unable to pursue alternative employment due to campus placement rules, while others declined confirmed offers from other employers in light of Wipro’s assurances.

The union also flagged concerns about a lack of transparency, alleging that while some candidates remained in an indefinite waiting period, Wipro continued onboarding others.

NITES said the issue is pan-India in nature, with more than 250 affected candidates approaching the union from different states.

The employee body has urged the ministry to seek an explanation from Wipro, direct the company to issue clear written decisions to all affected candidates, and either confirm onboarding timelines within a defined period or provide “reasoned closure so that their careers are not kept in indefinite suspension.”
Speaking to AIM, one affected fresher described repeated postponements after receiving an on-campus LoI.
“Initially, we were told that joining would happen around September 2025. Later, this was shifted to December. When December passed, we followed up again, and now we are being told that the joining may happen around April or May,” the candidate said.

“This constant delay has left us uncertain about our careers. We feel completely stagnated, as we are unable to move forward professionally,” he added.

Wipro did not respond to queries AIM sent on the matter.

However, its leadership has spoken about fresher hiring and headcount trends in recent earnings calls.

During the company’s Q2 earnings call, chief human resources officer Saurabh Govil said net headcount had increased and that Wipro had onboarded freshers from campuses, adding that hiring would continue based on demand and strong bookings.
In the Q3 call, CEO Srini Pallia said headcount growth was driven by acquisitions and large deal ramp-ups, and that there were no supply-side hiring challenges.

Govil also said campus recruitment during one quarter was deliberately slowed, with around 400 fresh NextGen associates onboarded, and that campus hiring would be ramped up in subsequent quarters, with total fresher hiring for the fiscal year expected to be around 7,500 to 8,000.

The post Employee Union Flags Fresher Onboarding Delays at Wipro, Seeks Labour Ministry Intervention appeared first on Analytics India Magazine.

LTIMindtree Q3 Profit Falls 12% to ₹960 Crore on Labour Code One-Off

LTIMindtree has reported a 12% year-on-year (YoY) decline in consolidated net profit for the third quarter of FY26, primarily due to a one-time charge related to the implementation of India’s new labour codes.

Net profit for the quarter stood at ₹959.6 crore, compared with ₹1,086.7 crore in the same period last year.
On a quarter-on-quarter basis, profit fell 30.5% from ₹1,381.2 crore in Q2FY26.

The drop was largely driven by a one-time labour code impact of ₹590.3 crore at the EBIT level, which translated into a ₹441.8 crore hit to net profit, the company said in its earnings release.

Excluding this exceptional cost, net profit rose 29% YoY to ₹1,401.3 crore and increased 1.5% sequentially. This indicates an improvement in underlying operating performance.
Venu Lambu, chief executive officer and managing director at LTIMindtree, said the quarterly performance reflects the impact of the company’s strategic AI pivot, continued success in large deals, and operational excellence. These are further supported by the company’s proactive efforts to build a more resilient and balanced portfolio.

“This marks our third consecutive quarter of 2%+ growth, highlighting our disciplined execution, deep tech-domain expertise, and differentiated AI-led offerings.”

Revenue from operations rose 11.6% YoY to ₹10,781 crore in the December quarter, up from ₹9,660.9 crore a year earlier. Sequentially, revenue grew 3.7% from ₹10,394.3 crore.

At the operating level, EBIT rose 5.4% quarter-on-quarter, while the operating margin expanded by 20 basis points to 16.1%, despite the one-time cost impact.
LTIMindtree said it had 746 active clients as of December 31 last year.
The number of ₹415-crore-plus ($5 million+) clients increased by 10 YoY to 162, ₹830-crore-plus ($10 million+) clients rose by seven to 97, and ₹1,660-crore-plus ($20 million+) clients increased by eight to 47.

The company’s workforce stood at 87,958 employees, following a net addition of 1,511 professionals in Q3.
Trailing 12-month attrition was 13.8%, while utilisation excluding trainees was 86.9%.

On the deal front, LTIMindtree highlighted five major wins. This includes a ₹1,285-crore ($155 million) deal with a leading US insurance and financial services company for a multi-year engagement. The deal covers application management, infrastructure operations, and end-user services, and will be driven by an AI-led delivery model.
It also announced multiple multi-year transformation deals across manufacturing, financial services, technology, and consumer sectors.

On January 16, the company announced it had been awarded the Insight 2.0 project by the Central Board of Direct Taxes (CBDT) to build an AI‑powered programme to modernise India’s national tax analytics platform.

Valued at around ₹3,000 crore, it said the seven-year mandate reinforces LTIMindtree’s leadership in enabling digital transformation through advanced digital architecture and data analytics to deliver real-time insights for policymakers.

The post LTIMindtree Q3 Profit Falls 12% to ₹960 Crore on Labour Code One-Off appeared first on Analytics India Magazine.

Tredence Introduces Agentic Commerce Accelerators for Retailers

How is Tredence Turning Professional Milestones into Family CelebrationsHow is Tredence Turning Professional Milestones into Family Celebrations

Tredence, on January 12, announced the launch of its Agentic Commerce solution accelerators, aimed at helping retailers design and deploy agent-driven shopping experiences faster as enterprises increasingly adopt generative AI across the customer journey.

The company said the enterprise-grade accelerators can cut time-to-value by up to 60% and act as configurable starting points for retailers to move towards mission-based shopping across omnichannel touchpoints.

“Commerce today is turning towards agentic systems that can sense, reason and act with velocity and accuracy,” said Sumit Mehra, chief technology officer and co-founder at Tredence. “The next phase of commerce will be defined by how intelligence is architected into the end-to-end shopper journey.”

The Agentic Commerce framework is structured as a system of agents that interprets shopper intent and orchestrates personalised interactions across the shopping lifecycle. The accelerators include customer intelligence, contextual search, content generation, shopper concierge and customer engagement agents spanning discovery, decision-making and post-purchase engagement.

Tredence is launching the accelerators initially on Google Cloud, with the first public showcase planned at the National Retail Federation’s annual retail event in New York. The company said the accelerators can be deployed across major cloud and data platforms depending on a retailer’s enterprise architecture, security and compliance needs.

“True transformation happens when advanced technology meets the right implementation partner,” said Jose Gomes, vice president of retail and consumer goods at Google Cloud. “Tredence translates our AI capabilities into retail-ready, multi-agent systems that deliver measurable business outcomes from day one.”

The company also cited an early deployment with Thorne, which has launched Taia, an AI-powered wellness advisor designed to provide personalised, evidence-based guidance while supporting product decisions.

The post Tredence Introduces Agentic Commerce Accelerators for Retailers appeared first on Analytics India Magazine.

Blackstone Confirms Milestone-Linked Investment in Neysa, Values Startup at $300 million: Reports

US alternative asset manager Blackstone has finalised a structured investment deal with Mumbai-based AI cloud infrastructure startup Neysa. The transaction could enable the firm to eventually acquire a controlling stake, The Economic Times reported.

The deal involves Blackstone investing $50-75 million in an initial tranche for a significant minority holding, with contractual rights to increase its stake to a majority if Neysa meets specific business milestones, according to the report. The transaction values Neysa at around $300 million, though the exact equity being acquired in the first phase was not disclosed.

The investment is expected to be among the largest fundraises by an AI company in India and marks a strategic move for Blackstone, which is increasingly bullish on data centres and AI-led infrastructure globally. The firm already has exposure to Indian data centres through Lumina CloudInfra, its joint venture with Panchshil Realty.

Unlike conventional venture capital rounds, the investment has been structured as a milestone-linked, control-oriented deal, reflecting the heavy capital requirements and long investment cycles associated with AI infrastructure businesses.

Neysa’s existing investors include Z47 (formerly Matrix Partners India), Nexus Venture Partners, Blume Ventures and Japanese telecommunications holding company NTT. The company has raised about $50 million across two rounds so far and was last valued at $120-130 million.

The report added that SoftBank had also held talks for a potential growth investment, but it is not part of the current transaction.

Founded in 2023 by entrepreneur Sharad Sanghi and former Netmagic executive Anindya Das, Neysa provides GPU-led cloud compute capacity and software platforms that enable enterprises, startups and government clients to build, deploy and manage large-scale artificial intelligence applications.

India’s data centre sector is projected to attract more than $50 billion in investments over the next five to seven years, driven by global hyperscalers and large domestic conglomerates, alongside policy support from the Centre and several state governments.

The post Blackstone Confirms Milestone-Linked Investment in Neysa, Values Startup at $300 million: Reports appeared first on Analytics India Magazine.

Baidu’s Apollo Go & AutoGo Launch Fully Autonomous Ride-Hailing in Abu Dhabi

Baidu’s autonomous ride-hailing service, Apollo Go and UAE-based AutoGo, owned by K2, have launched a fully autonomous commercial ride-hailing service in Abu Dhabi. The service is available via the AutoGo app.

The launch follows the partners securing a fully driverless commercial permit in mid-November 2025.

The initial operations cover Yas Island, which has been designated as a permitted zone for fully driverless operations. The companies said the service will expand in phases across Abu Dhabi.

The next phase will include Al Reem Island, Al Maryah Island, and Saadiyat Island. The partners said they will add more areas over time. The long-term plan is to operate across the wider Abu Dhabi emirate and deploy hundreds of vehicles by 2026.

Baidu said the collaboration with AutoGo began in March 2025. The partners announced plans then to build Abu Dhabi’s largest fully driverless fleet. By mid-November 2025, they secured one of the first permits for fully driverless commercial operations in the emirate.

Users can now download the app and request a ride, and the vehicles will operate without a human driver.

“This speed of execution highlights the technical readiness of Apollo Go, the strong operational capabilities of our partnership, and the steadfast support of local regulatory bodies,” said Liang Zhang, managing director of EMEA at Baidu Apollo.

AutoGo said the launch marks a shift from testing to public deployment. The company said it plans to expand services across key districts in phases. The partners said they will continue scaling the service to reach more users. They also said the deployment aligns with Abu Dhabi’s broader smart city goals.

“AutoGo’s transition to live robotaxi operations marks an important milestone in Abu Dhabi’s autonomous mobility journey,” said Sean Teo, managing director of K2. “Launching the service at the start of the year reflects our focus on execution and long-term value creation.”

“By introducing robotaxi services in real urban environments and scaling across key districts, we are moving decisively from development to deployment—delivering autonomy that is practical, safe, and ready for everyday use,” he added.

The company said Apollo Go has logged more than 240 million autonomous kilometres globally. More than 140 million kilometres were completed in fully driverless mode.

Apollo Go operates in 22 cities worldwide, according to the company. Its weekly ride count has surpassed 2.5 lakh. The service has completed more than 17 million cumulative rides as of October 31, 2025.

It has also previously partnered with Uber and CAR Inc in China.

The post Baidu’s Apollo Go & AutoGo Launch Fully Autonomous Ride-Hailing in Abu Dhabi appeared first on Analytics India Magazine.

Sequoia Breaks Ranks to Back Anthropic in $25 Bn Mega Round: Report

In a head-turning move, Sequoia Capital is set to join Anthropic’s cap table in a $25-billion funding round that will also see participation from Singapore’s GIC and US investor Coatue, the Financial Times reported. The investment would value the artificial intelligence startup at $350 billion—more than double its $170 billion valuation just four months ago.

Sequoia’s participation marks a notable shift from its traditional strategy. Venture capital firms typically avoid backing direct competitors, yet Sequoia already holds stakes in OpenAI and Elon Musk’s xAI, both rivals to Anthropic. Its investment in xAI, however, is widely seen less as a bet against OpenAI and more as an extension of its long-standing relationship with Musk. Sequoia also backed X when Musk acquired Twitter and rebranded the platform.

The $25 billion figure includes earlier commitments, with GIC and Coatue each planning to invest $1.5 billion. In late 2025, Microsoft and Nvidia pledged up to $15 billion to the company, while Anthropic also raised $13 billion in a Series F round in September last year, led by Fidelity, ICONIQ, and Lightspeed. The company plans to deploy the capital to develop more advanced AI systems and expand its technical infrastructure.

Anthropic reported sharp financial growth through 2025, with annualised revenue rising from $1 billion at the start of the year to $9 billion by December. The surge followed the success of its Claude chatbot, new tools for software developers, and the launch of specialised AI products for healthcare and financial services.

The company is also said to be preparing for a potential initial public offering, having consulted legal and financial advisers about a possible listing.

The funding round is expected to close in the coming weeks, and would rank among the largest private investments ever made in the technology sector.

The investment also highlights how funding is concentrated around a handful of leading AI companies, as soaring computing costs and fierce competition push investors towards scale players with proven products and revenues.

The post Sequoia Breaks Ranks to Back Anthropic in $25 Bn Mega Round: Report appeared first on Analytics India Magazine.

Top Indian IT Firms Added Only 17 Employees in Last 9 Months

Indian IT’s Goal to Hire 1 Lakh FreshersIndian IT’s Goal to Hire 1 Lakh Freshers

India’s top five IT services firms have added just 17 net employees in the first nine months of FY26, marking an almost absolute freeze in hiring, according to a report by The Times of India. The figure is a steep fall from the 17,764 net hires recorded in the same period last year.

The sharp slowdown reflects weak global demand, lower discretionary spending by clients, and the growing use of AI-driven delivery models that reduce the need for fresh hiring. Most limited headcount additions during the period reportedly came from acquisitions rather than direct recruitment.

TCS was the biggest drag on overall numbers after cutting 25,816 jobs following a workforce reduction plan focused largely on mid-level and senior roles. Infosys, Wipro, HCLTech, and Tech Mahindra reported additions to their employee base. However, these gains were offset by TCS’s large cuts, leaving the combined total almost flat.

In the December quarter alone, the combined headcount of the five companies fell by 2,174 employees. The broader trend signals a shift by IT firms towards tighter cost control, higher productivity, and more selective hiring aimed at niche skills instead of mass recruitment.

TCS leadership also made it clear that its workforce trimming is not over yet. The company has now signalled that more exits are coming in the March quarter as it moves towards a planned 2% reduction in headcount.

The post Top Indian IT Firms Added Only 17 Employees in Last 9 Months appeared first on Analytics India Magazine.

Why Postman CTO Believes APIs will Define the Era of AI Agents

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 Why Postman CTO Believes APIs will Define the Era of AI Agents appeared first on Analytics India Magazine.