← All digests

2026-05-15

India AI Digest — Friday, May 15, 2026

  • The Press Information Bureau published an AI-on-DPI feature on May 13 putting Banking BHASHINI, the Unified Lending Interface, MuleHunter.AI, and the RBI Regulatory Sandbox at the centre of the next phase of financial inclusion, with an unlockable USD 130–170 billion MSME credit-gap framing.
  • IBM and IndiaAI released a joint study on May 13 projecting USD 500-billion-plus AI contribution to India's GDP by 2030, surveying ~1,500 Indian executives; 72% of organisations say they trail global peers and 85% remain in pilot stage.
  • Uber CEO Dara Khosrowshahi announced in Ahmedabad on May 13 that Uber will build its first India data centre with the Adani Group; the facility is expected to go live later in 2026, with capacity, capex, and operating-entity details undisclosed.
  • Bengaluru-based HrdWyr closed a $13M Series A on May 12–13 led by Ideaspring Capital, with Singularity AMC, Avatar Growth Capital, and Persistent Systems participating, to build AI-native System-on-Chip products for edge and physical-AI workloads.
  • Google previewed Gemini Intelligence at The Android Show: I/O Edition on May 12 — an OS-layer agent moving across apps with on-screen context — alongside the Googlebook laptop category co-developed with five OEMs; rollout begins this summer on Samsung Galaxy and Pixel.
  • Silicon Road Ventures launched a SEBI-approved Category II AIF on May 14 with a target corpus of ₹150 crore for B2B Agentic-AI commerce-tech startups, led in India by Ajay Mahajan (ex-CARE Ratings MD & CEO).
  • Position movements: sectoral_maturity +1 (India fintech, mag 1, PIB framing); regulatory_clarity +1 (India lending, mag 1); compute_infrastructure +1 (India DC anchor demand, mag 2); domestic_capability +1 (India AI-native silicon, mag 1); consumer_adoption_depth +1 (India Android-base AI distribution, mag 2).

POLICY · FINTECH · INDIC LANGUAGE · May 13, 2026

PIB packages AI-on-DPI as the lever for the next phase of financial inclusion

The Press Information Bureau published a feature article on May 13, 2026 — "AI-Powered Financial Inclusion in India" — placing AI integrated with Digital Public Infrastructure at the centre of the country's next inclusion phase. The article names four operational primitives: Banking BHASHINI for multilingual banking, the Unified Lending Interface for credit, MuleHunter.AI (developed by the Reserve Bank Innovation Hub) for fraud control, and the RBI Regulatory Sandbox for live experimentation. It frames an unlockable MSME credit gap of USD 130–170 billion under AI-driven cash-flow and alternative-data underwriting that reaches segments currently below banks' cost-to-serve.

What this means. A PIB feature is not a policy instrument; it is government framing. The instrument-level work is being done in the agencies — RBI for ULI and MuleHunter.AI, DIBD/MeitY for Banking BHASHINI, the sandbox cohorts in the regulator. What the press note adds is the first PIB-grade consolidation of these primitives under one narrative: India's bet is use-case-led, DPI-anchored, and sectoral-regulator-supervised — not a frontier-model bet.

The MSME credit-gap framing is the headline metric, and it is the move worth holding. The argument is an underwriting unit-economics argument dressed as financial inclusion. AI-driven assessment of cash-flow and alternative-data signals lowers cost-to-serve below ticket sizes that have been uneconomic for banks. The USD 130–170 billion is the unlockable ceiling under that frame, not a projection of disbursals. The substance test is downstream — what ULI-on-AI cohorts disburse, what MuleHunter.AI cuts in fraud losses, what Banking BHASHINI does to per-customer onboarding cost. None of those numbers are in the article. Treat the framing as a policy signal about the model India will pursue, not as a forecast.

India angle. Three reads worth holding apart. For BFSI builders, the PIB note is the clearest articulation that the regulator-affiliated stack — ULI, MuleHunter.AI, Banking BHASHINI, RBI Sandbox — is the substrate to build on, not around. Vendors selling parallel infrastructure into Indian banks should expect a longer conversation than vendors that integrate. For the foundation-model cohort (Sarvam, AI4Bharat, Krutrim, others), the framing signals that the path to public-sector and BFSI revenue runs through Indic capability layered on DPI rather than through frontier-model parameter chases — a continuation of the pattern Banking BHASHINI exemplifies after the DIBD–RBI MoU of February 24, 2026. For policy watchers, the use-case-led framing is the live alternative to a horizontal AI law; the November 2025 India AI Governance Guidelines' sectoral-regulator approach gets a PIB-level statement of intent. The horizontal-law debate is not closed, but the centre of gravity in official communication is moving toward sectoral execution.

Behind the news. The PIB framing sits on a multi-quarter arc: Banking BHASHINI was launched via the DIBD–RBI MoU on February 24, 2026; MuleHunter.AI has been an RBIH-developed primitive in the public discourse since 2024; ULI began field-scale rollout under the RBI in 2024. What is new is the explicit packaging into one official narrative line. The DPI-plus-AI thesis has been operating-level visible since the IndiaAI Mission's 2024 framing of public-good AI; this is the first time it has been articulated as a single inclusion thesis with all four primitives named together.

What to watch. First MuleHunter.AI operational metric — bank-onboarding count or quarterly fraud-loss attribution — published by RBIH. Also: any ULI-on-AI cohort disbursal number from RBI's next monetary-policy or developmental-update statement. A horizontal AI law would force a re-read of this framing; a continued instrument-by-instrument rollout would confirm it.

Source: Press Information Bureau, May 13, 2026. → link

Confidence: high — PIB article retrieved as primary source; MSME credit-gap figure sourced to the article; operational metrics on cited primitives not in the article and remain to be measured.

RESEARCH · STRATEGY · ENTERPRISE · May 13, 2026

IBM and IndiaAI put a $500 billion 2030 number on the AI economy, with an execution gap

IBM Institute for Business Value and IndiaAI released a joint study on May 13, 2026 — "From promise to power: How AI is redefining India's economic future" — projecting that AI could add more than USD 500 billion to India's GDP by 2030. The study surveyed roughly 1,500 Indian executives across industries and leadership levels. Headline findings: 73% of Indian executives expect India to emerge as a leading global AI nation by 2030; 72% concede their organisations trail global peers in adoption; 15% are scaling AI cross-functionally while 85% remain pilot-stage; 57% cite uneven data quality and 77% cite the lack of accessible, affordable, secure cloud infrastructure as the binding constraints.

What this means. The headline number is a sizing exercise, not a forecast. USD 500 billion by 2030 is what survey respondents and IBM-IndiaAI's modelling layer count as AI's potential contribution to Indian GDP if execution closes; the report itself names the execution gap as the load-bearing question. The interesting datapoint is the 85% pilot-stage figure paired with the 72% trail-global-peers admission. The most-quoted survey-of-Indian-executives finding of the last two years has typically been some variant of "AI is a top board priority" — which is consistent with everything, including pilots that never reach production. The 85/72 pairing is the more honest read: enthusiasm is high; deployed scale is not, and the executives surveyed know it.

The infrastructure number is the one to dwell on. 77% citing lack of accessible, affordable, secure cloud infrastructure is the explicit corporate-buyer version of the constraint that the IndiaAI Mission's GPU subsidy has been pointing at from the supply side. Two different framings of the same gap: founders say compute is too expensive; CIOs say cloud-AI infrastructure they can deploy on regulated workloads does not exist at price. Either framing converges on the same operational story — Indian enterprise AI is a deployment story, not a model-availability story, and the bottleneck is the layer between cloud and use case.

The talent number is the under-discussed one. The report says only ~30% of employees have the AI literacy businesses need today, against a 2030 requirement closer to 57% — a 350-million-person delta. That is the IndiaAI Skilling Mission's de facto TAM, and it is also a hiring problem at the SI / GCC layer that no single training programme will close in five years.

India angle. Two reads, both useful. For Indian SIs and GCCs (TCS, Infosys, Wipro, HCL, Cognizant captives), the survey is the customer's confession that they cannot scale AI in-house — which is the SI sale, except the offering has to be reframed from advisory to deployed outcomes against pilot fatigue. Catalog churn at SI scale, with the AI-readiness consult repackaged as concrete production deployment SLAs. For the Indian AI-native vendor cohort (data, MLOps, evaluation, agent orchestration), the 77% infrastructure constraint is the addressable market; the question is whether the Indian buyer wants a domestic stack or a localised version of a US stack, and the answer is currently mixed. For policy, the talent gap is the next clearly-quantified target after compute — the IndiaAI Future Skills programme already exists; the question is whether the funding line scales with the gap.

Behind the news. Industry-funded sizing studies of Indian AI markets have been a fixture for at least three years — Zinnov-Z47-OpenAI's India AI Adoption Edge 2026 dropped the same week with a different framing. What is novel here is the IndiaAI co-imprimatur on a number. Government-affiliated sizing of the opportunity creates an anchor that future Mission outlay debates can reference. Whether subsequent IndiaAI Mission communications cite this figure or push past it will indicate how official Delhi reads the report.

What to watch. Whether IndiaAI's next Mission update cites the USD 500B figure as the official sizing or moves to its own. Also: the IBM-IndiaAI follow-on methodology disclosure — the survey-derived components vs. the modelling components of the USD 500B number are not separable from the press summary, and the substance of the projection sits there.

What this is not. Not a forecast. The USD 500B is an unlockable-potential number under a set of assumptions about adoption, productivity, and execution that the report itself flags as not-yet-evidenced. Treating the figure as a base case rather than a stretch case misreads it. The companion datapoint — 85% pilot-stage — is the contemporaneous reality the number is a counterfactual to.

Source: IBM Newsroom India, May 13, 2026; ANI / Business Standard / Daily Pioneer, May 13–14, 2026. → link

Confidence: medium — top-line and survey statistics consistent across IBM India newsroom and multiple secondary outlets; full methodology and IBM IBV report PDF not retrieved as primary source.

COMPUTE · INFRA · STRATEGY · May 13, 2026

Uber and Adani Group to set up Uber's first India data centre

Uber CEO Dara Khosrowshahi announced on May 13, 2026 in Ahmedabad, following a meeting with Adani Group chairman Gautam Adani, that Uber will set up its first India data centre in partnership with the Adani Group. The facility is expected to go live later in 2026 and to support Uber's global technology operations — ride matching, dynamic pricing, navigation routing, demand forecasting, safety monitoring — run from India. Capacity, capex, location, and the operating Adani entity (AdaniConneX, the group's 50:50 hyperscale JV with EdgeConneX, is the likely vehicle) were not disclosed at the announcement.

What this means. A global consumer-tech company committing to its first India data centre with a domestic infrastructure partner adds anchor demand to the Indian colocation pipeline. The 6-to-18-month arc is foreign-platform compute localisation — DPDP-era data-residency posture, the IndiaAI Mission compute push, and the steady migration of regulated workloads onto in-country infrastructure. Uber has not framed this as a sovereign-compute play; the stated rationale is operating global products from India, which reads as engineering-region consolidation rather than residency compliance. Both readings are present in the announcement; the operational reality will turn on the workload mix the facility ends up serving.

The Adani side is the part most easily under-read. AdaniConneX is targeting roughly 1 GW of installed DC capacity by 2030. An anchor tenant of Uber's profile strengthens the build-out thesis without disclosing what fraction of capacity Uber occupies. The substance test is whether the facility appears in AdaniConneX's next capacity-update or in Adani Enterprises' financial disclosures with site-specific capex, and whether a second similar anchor tenant follows. The announcement is source-conditional in the strict sense that both parties are the only entities asserting the timeline; no third-party verification of capacity or capex is currently available.

India angle. Three reads. For the data-centre-operator cohort (CtrlS, NTT, AdaniConneX, Yotta, STT), an Uber-anchored Adani facility is the kind of name-brand anchor demand the segment has been chasing — useful as a competitive marker even for non-Adani operators. For Indian enterprise buyers, more in-country capacity from name-brand operators broadens the menu of residency-bound compute, which matters most for BFSI and government workloads. For the engineering-region narrative, Uber framing the facility as supporting global tech operations from India is a small but real signal that India is moving from back-office region to production region for some platforms — the same arc TCS, Infosys, Cognizant and several captives have been on, but with a US consumer-tech company as the example.

Behind the news. Fits the broader May 2026 India-compute storyline — Cabinet semicon nods, IndiaAI Mission GPU subsidy, and foreign-platform residency posture have been the recurring elements. Foreign platforms localising compute in India after DPDP and during the IndiaAI Mission's compute push is now a pattern rather than a one-off.

What to watch. First AdaniConneX or Adani Enterprises disclosure that names the Uber facility with capacity or capex specifics; third-party verification of the go-live date in H2 2026; any second comparable anchor-tenant announcement on AdaniConneX, which would convert today's signal from one-off to repeatable pattern.

Source: Business Standard, May 13, 2026; The Tech Portal, May 13, 2026; Indian Infrastructure / News9, May 14, 2026. → link

Confidence: medium — announcement and the two named principals confirmed by multiple Indian outlets; capacity, capex, location, and operating-entity details not disclosed; go-live timing self-reported by the announcing parties.

SEMICONDUCTOR · COMPUTE · FUNDING · May 12, 2026

HrdWyr raises $13M Series A to build AI-native System-on-Chip products for edge

Bengaluru-based HrdWyr closed a $13 million Series A funding round announced on May 12, 2026, led by Ideaspring Capital, with participation from Singularity AMC, Avatar Growth Capital, and existing investor Persistent Systems. Founded in 2023 by Ramamurthy Sivakumar and Guruswamy Ganesh, the company describes itself as a fabless semiconductor product company building AI-native System-on-Chip (AISoC) products designed "from first principles" for data-intensive edge and physical-AI environments. The capital will accelerate AISoC development and customer engagements across global markets.

What this means. A $13M Series A is mid-stage capital for a fabless silicon company — enough to push from architecture and emulation toward tape-out and first samples on a leading-node process, not enough to absorb a tape-out miss or run a second-product spin. The decisive question for any fabless startup at this stage is the customer-pull story: who is the design-in commitment with, what is the production-volume target, and what is the timeline from first samples to revenue. None of those are in the public summary of this round. Treat the round as a credibility marker — Ideaspring is a deep-tech-focused Indian fund, Persistent Systems' continued participation reads as a strategic believer, and Avatar Growth and Singularity AMC fill out a syndicate of operators rather than tourists — but treat the AISoC product as forthcoming, not shipping.

The "AI-native" framing is the marketing claim that needs the most discipline. The serious version of the claim is that the chip's memory-compute organisation, dataflow, and instruction set are designed around transformer-class and physical-AI workloads from the outset, rather than CPU- or GPU-derived. The unserious version is a marketing wrapper around a generic NPU. The difference shows up in disclosed benchmarks, programmer-facing documentation, and external silicon reviews when first parts ship. None of those exist publicly today. The substance test is what HrdWyr publishes at tape-out and sampling, not at funding announcement.

Edge inference is the bet that has the cleanest unit-economics story for an Indian fabless attempt — lower power budgets, tighter latency, and physical-AI use cases (robotics, vehicles, industrial sensing) where on-device inference is the architectural answer rather than a cost compromise. Cloud-class training silicon would be a five-year, hundred-million-dollar bet against entrenched incumbents; edge inference is a tractable foothold for a $13M Series A.

India angle. Three reads. For the India-Semiconductor-Mission narrative, HrdWyr is a fabless product-company datapoint that complements the fab-and-OSAT-heavy ISM project list. ISM's headline projects have been Tata-PSMC, Micron-Sanand, CG Power, Kaynes, and the more recent Gujarat units; product companies that design the silicon those fabs would eventually fabricate have been thinner. HrdWyr (and the small handful of peers — Mindgrove, InCore, Calligo) is the layer above. For Indian VC, deep-tech rounds at $13M have been rare enough that each one is a forcing function for syndicate composition; the Ideaspring–Singularity–Avatar–Persistent mix is closer to a deep-tech-syndicate template than to a SaaS round. For the physical-AI cohort in India (robotics, drones, agritech, edge-vision), domestic edge-inference silicon eventually translates into BOM-cost flexibility — but that benefit lands when parts ship, not at funding.

Behind the news. The Indian fabless ecosystem has been building visibility through 2024-2026 — Mindgrove's MIND-CT and InCore Semiconductors' RISC-V cores are the most-cited prior datapoints — but consistent Series A capital into product fabless has lagged the ISM's downstream-fab announcements. HrdWyr's round is on the same trend line as recent Sankalp Semiconductor / KPIT moves and continues a slow normalisation of fabless-product funding in India.

What to watch. First public AISoC product spec — node, target wattage, ML benchmarks against named comparables (Qualcomm Cloud AI 100 / NXP / Hailo / Tenstorrent equivalents). First named design-in customer. Tape-out timeline if disclosed. Whether the syndicate adds a strategic chipmaker or hyperscaler in the next round.

Source: Entrackr, May 12, 2026; BusinessToday, May 12, 2026; Deccan Herald, May 14, 2026; YourStory. → link

Confidence: medium — round size, syndicate, and founder names corroborated across four Indian outlets; product specifics, design-in customers, and tape-out timeline not disclosed.

MODEL RELEASE · CONSUMER · AGENTS · May 12, 2026

Google previews Gemini Intelligence for Android; introduces Googlebook laptop category

At The Android Show: I/O Edition on May 12, 2026, Google previewed Gemini Intelligence — described as an OS-layer agent that moves across Android apps, reads on-screen context, and completes multi-step tasks. The release bundle included Gemini-in-Chrome with auto-browse, a Gboard "Rambler" voice-input upgrade, AI-generated homescreen widgets, and Googlebook — a co-developed laptop category with Acer, Asus, Dell, HP, and Lenovo built ground-up for Gemini Intelligence. Google said agentic features begin rolling out on the latest Samsung Galaxy and Google Pixel devices this summer, with Googlebooks shipping in the fall.

What this means. Two distinct stories. Gemini Intelligence is the larger of the two. An OS-layer agent that reads on-screen context and operates across apps is a different surface from in-app assistants — it folds the assistant into the platform, not the app. The Android distribution base makes this a consumer-AI distribution event of the kind only the platform incumbents (Google, Apple, Samsung at the OEM layer, Microsoft on Windows) can do. The framing of "racing Apple's AI reboot" is competitive positioning, and the substance test is the rollout cohort: which Galaxy and Pixel models, which capabilities ship at GA versus stay behind on-device-model thresholds, and how on-screen-context interacts with app-level permissions.

The Googlebook unveil is the more strategic-but-thinner story. A Chromebook-adjacent laptop category co-developed with five OEMs and architected around Gemini Intelligence is Google extending the OS-as-AI-surface story off-phone. It is also a defensive move: Microsoft's Copilot+ PC push and Apple's M-series AI features have been building a laptop-class AI tier that Chromebook OEMs have lacked an answer to. Hardware specifics, pricing, and on-device-model story are not disclosed. Treat Googlebook as the announcement of a category, not as a shipping product cohort.

The Indic-coverage question is the gap in today's announcement. An OS-layer assistant that reads on-screen context becomes meaningful for the majority of Android users in India only if Indic input, output, and on-screen recognition cover Indic scripts and code-switched registers (Hinglish, Tanglish, Manglish). The release materials do not detail Indic coverage; until they do, the India-relevance read is conditional on a follow-on disclosure.

India angle. Three implications. India has Android's largest installed base; an OS-layer agent shipping to Galaxy and Pixel is the single largest consumer-AI distribution surface that touches India this year, even on a partial rollout. The interaction with DPDP becomes live the moment the agent ships to Indian users — on-screen context-reading is a personal-data primitive, and the consent UX and processing-locus details have not been disclosed for the Indian market. For the Indic-NLP cohort (AI4Bharat, Sarvam, Bhashini), Gemini Intelligence becomes both a benchmark and an integration question — whether Indic capability shows up as Google's native model coverage, as a Bhashini-style backend integration, or not at all in v1 will signal where the platform incumbent thinks the Indic-model layer should sit.

Behind the news. The OS-layer agent move sits on a 12-18 month arc that includes Apple Intelligence, Microsoft Copilot+, and the Samsung-Google Galaxy AI integration of the past year. The announcement is Google's pre-Apple-WWDC competitive positioning for the Android base. The Indic-language question has been live since AI4Bharat began publishing IndicTrans and IndicBERT benchmarks; what is new is the OS-layer surface as the integration point.

What to watch. Samsung Galaxy and Pixel rollout cohort with model coverage when the summer rollout begins. Any Google or Bhashini-side disclosure on Indic-language coverage for Gemini Intelligence. Googlebook pricing, OEM mix, and Indic-typing support when it reaches retail.

What this is not. Not a Pixel-only feature. The Samsung Galaxy inclusion is the bigger India read, since Samsung's share of the Indian Android premium-mid tier dwarfs Pixel's. The India-distribution story depends on Samsung's rollout schedule, not Google's flagship cohort.

Source: Google blog (The Android Show: I/O Edition), May 12, 2026; TechCrunch, May 12, 2026; CNBC, May 12, 2026; Business Standard, May 13, 2026. → link

Confidence: medium — announcement details corroborated across multiple outlets and Google's own blog; Indic-language coverage details not disclosed; India rollout specifics not separated from global rollout in announcement materials.

FUNDING · ENTERPRISE · AGENTS · May 14, 2026

Silicon Road Ventures floats ₹150 crore SEBI-approved India AIF for agentic-AI commerce

Atlanta-headquartered early-stage venture capital firm Silicon Road Ventures announced on May 14, 2026 the launch of its India-focused fund, structured as a SEBI-approved Category II Alternative Investment Fund with a target corpus of ₹150 crore. The fund has achieved first close and has commenced active deployment. Investment thesis: early-stage startups building B2B agentic-AI solutions in commerce technology — multi-channel commerce, supply chain and logistics, fintech, consumer insights, and retail operations. Indian investing is led by Partner Ajay Mahajan, formerly MD & CEO of CARE Ratings and earlier in leadership roles at Bank of America, Yes Bank, UBS, and IDFC. SRV, founded in 2019 by Sid Mookerji, has previously backed over 30 startups in retail and enterprise technology.

What this means. A ₹150 crore fund is small-mid scale by Indian VC AIF standards — roughly USD 18 million at current rates — but the thesis specificity is the more interesting part. "B2B agentic-AI commerce" carves out a narrower mandate than "AI for India" generalist funds, and the partner profile is unusual for a fund of this size: a former rating-agency MD & CEO with three decades of banking leadership is not a typical first-cheque partner. Agentic commerce as a category covers two distinct buyer motions — pure-play AI vendors selling agent layers to existing commerce stacks, and commerce-tech incumbents embedding agentic features. Which of those two motions the fund mostly backs will determine whether it competes with horizontal AI funds or with retail-tech specialists.

The SEBI Category II AIF wrapper is the operating detail worth noting. India-domiciled Cat II AIFs invest predominantly in unlisted Indian companies, with capital from Indian and foreign LPs. For a US-based GP, this structure is the path to deploy into Indian early-stage at scale without the round-trip and FEMA complexity of foreign-fund onshore investments. The structure choice signals an intent to operate locally as an Indian investor, not as a US fund with India exposure. Whether SRV builds out an India team beyond the partner-level appointment is the next signal.

The thesis sits on a real demand wave. Indian retail and CPG buyers — Reliance Retail, DMart, Tata Trent, ITC, HUL, Marico, Dabur — have been increasing AI-adjacent spend through 2025-2026, with most deployments still in pilots that mirror the IBM-IndiaAI 85% pilot-stage finding from the same week. The agentic layer is the next-product wedge: where workflow automation in commerce has historically been rules-based, agent frameworks open multi-step task automation that targets the actual cost lines (call-centre, returns, supply-chain exception handling, merchandising decisions). The substance test is which Indian agentic-commerce startup actually books production deployment revenue at one of the named retail buyers — to date, the count is small.

India angle. Three reads. For Indian agentic-AI founders, an India-domiciled GP with a named sectoral thesis at the first-cheque size is the kind of capital that has been thin on the ground — most Indian AI capital is generalist, and most agentic-commerce demand sits with retailers and CPG buyers who require local-onshore sales motions. For Indian retail-tech incumbents (Capillary, Wooqer, Bizom, Increff), a US-backed Indian agentic-commerce wedge increases the risk that the agent layer gets built next to them rather than by them. For the broader Indian AI capital ecosystem, SRV joins a small set of foreign-GP India AIFs (Iron Pillar, Lightspeed India, Peak XV) that have crossed the structural-commitment threshold — a useful peer benchmark for the next foreign AI-focused fund weighing whether to onshore.

Behind the news. Foreign-VC India AIFs have been compounding since 2024 — Lightspeed and Peak XV's India structures are the older versions; newer entrants have tended toward thematic Cat II vehicles like this. Agentic AI as a thematic carve-out is a 2025-onwards entrant in Indian fund branding; the substance-vs-marketing question is open across the cohort. SRV's retail/enterprise pedigree gives the thesis a sectoral spine that pure-thematic agent funds typically lack.

What to watch. First disclosed portfolio company under the new fund, and whether the cheque size lands in the seed (₹3–8 crore) or pre-Series A (₹10–25 crore) band. Final close size if it exceeds the ₹150 crore target. Whether SRV makes additional India-team hires beyond Ajay Mahajan.

Source: Business Standard / ANI, May 14, 2026; The Tribune, May 14, 2026; Entrackr, May 14, 2026; Newskart. → link

Confidence: medium — fund structure, corpus target, SEBI category, and Indian-partner appointment corroborated across multiple Indian outlets; LP composition and first-close-size details not disclosed; portfolio deployment is asserted by the fund and not yet third-party-verified.