Request for Startups — India Edition

ThesisA model lab building genuinely capable multilingual reasoning models native to India's 22 official languages and dozens of scripts — not English models with a translation layer bolted on.

Why nowSubsidized national compute and sovereign-model financing have collapsed the capital barrier that made this impossible two years ago; the market is the second-largest internet population on earth, most of which the English-first frontier ignores.

TailwindIndiaAI Mission, Deep Tech FoF, Peak XV/Accel AI thesis.

ThesisSpeech-to-speech infrastructure (ASR, TTS, and full duplex voice agents) that works across code-switched, accented, low-resource Indian speech — the default interface for the next 500M users who will never type a query.

Why nowVoice is how Bharat actually computes; the data and model quality have finally crossed the usability threshold, and every vertical agent below needs this as a dependency.

TailwindAccel Atoms (voice AI explicitly), IndiaAI.

ThesisTesting, benchmarking, and safety-evaluation tooling for models deployed in Indian languages and regulated domains (finance, health, government), where global eval suites are blind.

Why nowEnterprises and the state are deploying AI faster than they can verify it; "does this model hallucinate in Marathi about a loan?" is an unanswered, fundable question.

TailwindIndiaAI responsible-AI mandate, enterprise demand.

ThesisInference optimization, model distillation, and serving infrastructure that makes frontier-class capability run cheaply on constrained or subsidized Indian compute.

Why nowIndia will run on borrowed and rationed GPUs for years; whoever makes a 7B model perform like a 70B one owns the cost curve for the whole ecosystem.

TailwindIndiaAI compute program, Peak XV infra thesis.

ThesisThe data company that builds high-quality, rights-cleared, India-specific training and evaluation corpora — legal, medical, agricultural, dialectal — as a product.

Why nowThe constraint on Indic models is not compute anymore, it's data; a neutral data layer becomes critical infrastructure everyone pays for.

TailwindIndiaAI datasets pillar.

ThesisTiny, efficient models and runtimes that deliver useful AI offline on low-end Android hardware with poor connectivity.

Why nowThe cloud-first assumption excludes most of India; edge inference turns a billion cheap devices into the distribution channel.

TailwindIndiaAI, consumer-hardware investor interest (Antler).

ThesisA repository of one-click, deployable integrations connecting internal business systems to the latest models, so any company can experiment, evaluate, and adopt AI against its proprietary data without a platform team.

Why nowEnterprise GenAI penetration is still below 10% — the bottleneck is plumbing, not appetite.

TailwindPeak XV developer-tooling thesis; named explicitly in Antler's NEXT100.

ThesisA technology-first insurer/broker that automates product, sales, underwriting, and claims — selling the outcome (a policy, a settled claim), not software to insurers.

Why now~35–40% of premiums are eaten by operational inefficiency and 10–15% by fraud; a low-cost AI-native insurer can finally reach the 700M+ under-insured lower-middle-income segment.

TailwindAntler (explicit NEXT100 thesis), YC services-as-software RFS.

ThesisAn AI-native firm that does the books, files the returns, and runs the audit for India's millions of SMBs and the global mid-market — replacing the service, not assisting the accountant.

Why nowThis is the canonical "services revenue >> software revenue" wedge, and Indian compliance complexity is a moat against generic global players.

TailwindYC RFS (accounting/tax/audit named), Peak XV.

ThesisEnd-to-end handling of high-volume legal work — contracts, notices, registrations, small-claims, compliance filings — priced per outcome.

Why nowIndia's justice and contracting systems are bottlenecked by cost and access; a vernacular AI legal service expands the market rather than splitting it.

TailwindServices-as-software thesis; DPI document rails.

ThesisA creative or performance agency with software economics — selling finished output at high value with near-zero marginal cost, decoupling revenue from headcount.

Why nowGeneration quality is now good enough that the constraint is taste and orchestration, both of which India has in surplus relative to its cost base.

TailwindYC RFS (AI-native agencies), global demand + Indian cost arbitrage.

ThesisAn IT-services firm built AI-native from day one, selling intelligence as an outcome (autonomous analysis, decisions, optimization) rather than software licenses and bodies on contract.

Why nowAgentic AI is dissolving the per-seat, per-hour model that India's $250B IT industry was built on; the incumbents are structurally slow to cannibalize themselves.

TailwindTVS Capital / Antler NEXT100 thesis; deep Indian talent pool.

ThesisAutomate prior-auth, coding, claims, and billing for hospitals and insurers as a managed outcome.

Why nowHealthcare admin is pure friction, heavily outsourced, and therefore easy to replace wholesale with an AI-native operator.

TailwindYC RFS (healthcare admin named), Ayushman Bharat scale.

ThesisContinuous, autonomous regulatory compliance (KYC/AML, labour, environmental, sectoral) sold as an always-on service to manufacturers, fintechs, and exporters.

Why nowIndian regulatory surface area is vast and changing; "we keep you compliant" is a stickier promise than "we sell you a dashboard."

TailwindYC RFS (compliance), manufacturing/export tailwinds.

ThesisSource, screen, interview, and place candidates end-to-end as an outcome service — particularly for blue- and grey-collar hiring at scale.

Why nowIndia hires tens of millions annually through a broken, offline funnel; voice + vernacular AI can run the whole loop.

TailwindPeak XV (Alex/recruiting precedent), Accel.

ThesisA buyer's AI that filters, prioritizes, and negotiates inbound sales outreach based on the buyer's real needs — bringing efficiency to the buying side, not just selling.

Why nowAI has made outbound infinite and worthless; the scarce, valuable position is on the buyer's side of the table.

TailwindAntler NEXT100 (explicit), enterprise procurement spend.

ThesisVertical AI co-pilots that run sales, support, operations, and compliance for specific SMB categories (pharma distributors, auto dealers, clinics, kirana wholesalers) in Indian languages.

Why nowIndia's 60M SMBs are ~30% of GDP with sub-10% GenAI penetration and an urgent appetite for productivity; the moat is fine-tuning to language and industry nuance.

TailwindGoogle/Accel "AI agents for India's SMBs," Peak XV vertical SaaS.

ThesisA super-simple AI platform for micro-businesses — bookkeeping, inventory, digital marketing, customer retention — designed for the corner shop, not the enterprise.

Why nowThe informal economy is the largest underserved software market on earth; vernacular voice-first AI is the first interface it can actually use.

TailwindAntler NEXT100 (Haptik thesis), ONDC rails.

ThesisModels that don't just chat but take actions in a specific industry's systems — dispatching trucks, scheduling production, routing maintenance — for logistics, construction, maritime, energy, or waste.

Why nowAccel is explicitly hunting this; the winners pair model work with deep workflow integration competitors can't copy from a prompt.

TailwindAccel Atoms (named verticals), Deep Tech FoF.

ThesisA personal AI that cuts decision fatigue — organizing priorities, handling customer service, talent, and user interviews — so operators focus on high-impact work.

Why nowThe tools to integrate an owner's whole operational context now exist; the wedge is trust and daily habit.

TailwindAntler NEXT100 (explicit), consumer-prosumer crossover.

ThesisAgents that modernize the legacy BSS/OSS, ERP, and ticketing estates that Indian IT runs for the world — turning maintenance contracts into autonomous operations.

Why nowTrillions of dollars of enterprise legacy is the substrate; whoever builds the context graph and agent layer over it captures the modernization wave.

TailwindPeak XV/Accel enterprise AI, global outsourcing base.

ThesisTooling that automatically documents and maintains enterprise databases so AI agents can operate against them as "experienced data team members."

Why nowEvery enterprise agent fails on messy, undocumented data; this is the unglamorous layer that unblocks all the others.

TailwindAccel Atoms (Zingle precedent), dev-tooling thesis.

ThesisAutonomous agents that run ERP management tasks — reconciliations, master-data hygiene, workflow exceptions — inside the systems of record that large enterprises live in.

Why nowERP is where the money and the drudgery both sit; agentic automation here has immediate, measurable ROI.

TailwindAccel/Google Atoms 2026 (Dodge.ai precedent).

ThesisAI that hyper-personalizes skilling, motivation, and engagement per employee — tailored learning paths, real-time nudges, emotional support — instead of one-size-fits-all HR software.

Why nowAI makes individualized employee experience computable for the first time; India's young, churn-heavy workforce is the proving ground.

TailwindAntler NEXT100 (Darwinbox thesis), HR-tech market.

ThesisCredit models built on alternative, consented data (Account Aggregator flows, GST, UPI, telco, geospatial) to lend to the 400M+ Indians invisible to traditional bureaus.

Why nowThe Account Aggregator framework finally makes consented financial data portable; regulated risk-scoring is squarely in Peak XV's stated wheelhouse.

TailwindPeak XV (regulated fintech), DPI (Account Aggregator).

ThesisCredit, insurance, and working-capital products embedded natively into the open commerce network for its millions of small sellers.

Why nowONDC unbundles commerce from the platforms; financial services on top of open rails is a greenfield distribution channel.

TailwindONDC/DPI, fintech investor density.

ThesisA near-free AI financial advisor that brings genuine planning — not just a brokerage app — to first-time investors in their own language.

Why nowDemat accounts have exploded but advice has not; "everyone gets a near-free financial advisor" is a stated frontier thesis.

TailwindAntler NEXT100 (universal-expertise thesis), Groww/Peak XV precedent.

ThesisHumane, compliant, vernacular collections run by voice agents — recovering debt without the abuse and inefficiency of the current call-center model.

Why nowLending is scaling faster than recovery infrastructure; a compliant AI layer is both cheaper and lower-risk.

TailwindLending boom, RBI compliance pressure.

ThesisCheaper, faster rails for the world's largest remittance-receiving country and for the new wave of cross-border Indian businesses, routed through GIFT City / IFSC structures.

Why nowIndia receives >$100B in remittances and is building global companies from day one; the FX and settlement layer is still expensive and slow.

TailwindGIFT IFSC, Peak XV cross-border thesis.

ThesisAn AI CFO for small businesses that forecasts cash, automates invoicing/collections, and sources financing — closing the working-capital gap that kills MSMEs.

Why nowGST and digital invoicing created the data; the missing piece is the autonomous layer that acts on it.

TailwindDeep Tech FoF (fintech AIFs), MSME policy focus.

ThesisParametric, sensor-and-satellite-triggered insurance for farmers, gig workers, and small businesses exposed to heat, flood, and crop risk.

Why nowClimate volatility is rising and geospatial data is now legal to collect and commercialize in India; payouts can be instant and disputeless.

TailwindGeospatial liberalization, BIRAC/climate funds.

ThesisA fraud and identity layer purpose-built for an economy where AI agents will soon transact on humans' behalf over UPI.

Why nowAgentic commerce breaks the assumptions UPI fraud systems were built on; someone has to secure machine-to-machine payments.

TailwindDPI scale, Peak XV (Better Auth precedent).

ThesisAn AI-plus-human primary-care service that gives every Indian an always-available first point of contact — triage, chronic-disease management, follow-up — in their language.

Why nowIndia's doctor-to-patient ratio guarantees access can never be solved with humans alone; AI-human hybrids make quality primary care a basic right, not a privilege.

TailwindAntler NEXT100 (HealthifyMe thesis), Ayushman Bharat.

ThesisContinuous, personalized management of diabetes, hypertension, and obesity — the diseases that will define India's health burden — blending wearables, AI coaching, and clinicians.

Why nowIndia is the diabetes capital of the world; continuous data plus AI coaching changes outcomes at a cost the system can bear.

TailwindConsumer-health investors, wearable adoption.

ThesisPoint-of-care diagnostic AI (imaging, pathology, screening) deployable on cheap hardware by a frontline health worker, not a specialist.

Why nowModel quality now matches specialists on narrow tasks; the bottleneck is rugged, offline, regulator-cleared deployment.

TailwindBIRAC, IndiaAI, medical-device DLI interest.

ThesisA computational discovery platform aimed at diseases global pharma ignores but that India carries the burden of (TB, dengue, antimicrobial resistance).

Why nowThe BioE3 policy and BIRAC are funding bio-foundries; India can own the diseases the rest of the world deprioritizes.

TailwindBIRAC / BioE3, ANRF, RDI scheme.

ThesisAn AI research assistant that accelerates R&D across biology, chemistry, and materials — generating hypotheses, designing experiments, reading the literature.

Why nowThis is a stated frontier (Accel/Google backed a "co-scientist"); India's research institutions are large and under-tooled.

TailwindAccel Atoms (K-Dense precedent), ANRF.

ThesisAccessible, vernacular, AI-assisted mental-health support with clear clinical guardrails and human escalation.

Why nowIndia's mental-health gap is enormous and stigmatized; private, low-cost, language-native support reaches people clinics never will. (Build with rigorous safety design.)

TailwindHealthcare investors, digital-health funding.

ThesisProducts addressing the systematically ignored stages — reproductive health, maternal care, menopause — for India and emerging markets.

Why nowA globally recognized white space with strong demand and almost no quality vernacular product.

TailwindConsumer-health investors, BIRAC.

ThesisEngineered biology for high-value materials, enzymes, and bio-based chemicals — turning India's fermentation and agri base into a bioeconomy.

Why nowThe BioE3 (Biotechnology for Economy, Environment, Employment) policy is explicit national strategy; the inputs and talent are domestic.

TailwindBioE3, BIRAC, Deep Tech FoF.

ThesisA voice-first, vernacular advisor that gives 120M+ farmers crop-specific, hyper-local guidance on inputs, pests, irrigation, and price.

Why nowCheap edge AI plus voice plus satellite data finally makes per-farmer advice economic; the moat is data density and trust.

TailwindIndiaAI, RKVY-RAFTAAR, Ankur Capital.

ThesisAI vision plus robotics that identify individual weeds and pests in real time and treat one plant instead of blanket-spraying a field.

Why nowYC's headline 2026 RFS (cutting pesticide use ~90% while raising yields) applies even more forcefully where smallholder economics are tight and chemical overuse is poisoning soil and water.

TailwindYC RFS (agriculture), Accel (agri action models), iDEX-adjacent drones.

ThesisMicrobes, peptides, and RNA-based crop protection that displace synthetic pesticides and fertilizers.

Why nowBiology has caught up; regulatory and consumer pressure on glyphosate-class chemicals is mounting globally.

TailwindBioE3, BIRAC, climate funds.

ThesisAn AI-orchestrated supply chain that cuts the 30%+ post-harvest losses between farmgate and consumer through better demand sensing, cold chain, and routing.

Why nowIndia wastes a staggering share of what it grows; the data and logistics tooling to fix it now exist.

TailwindONDC (agri), logistics investors.

ThesisGene-edited resilient crops and economically viable controlled-environment agriculture for a heating, water-stressed subcontinent.

Why nowHeat and erratic monsoons are now the baseline; resilience is a survival input, not a premium.

TailwindBioE3, ANRF, Accel LeapTech.

ThesisSensor-and-AI-driven optimization for India's vast aquaculture and fisheries — feed, disease, water quality, and traceability for export.

Why nowSeafood is a major export with thin margins and high disease risk; maritime is a named Accel vertical and an under-built one.

TailwindAccel (maritime action models), export incentives.

ThesisMeasurement, reporting, and verification infrastructure that lets millions of small farmers actually access carbon and regenerative-agriculture markets.

Why nowGeospatial liberalization plus satellite AI makes credible, low-cost MRV possible; without it, smallholders are locked out of climate finance.

TailwindGeospatial policy, climate funds, RDI.

ThesisAI software that forecasts, balances, and trades across an increasingly distributed, solar-and-wind-heavy grid — the operating system for the energy transition.

Why nowIndia is adding renewables faster than the grid was designed for; AI-driven demand response and forecasting is the cheapest way to add effective capacity, especially as AI datacenters spike demand.

TailwindRDI scheme, energy-transition funds, Accel (energy vertical).

ThesisStorage built for Indian conditions — heat-tolerant, low-cost chemistries (sodium-ion, aluminium-air, flow) that don't depend on imported lithium supply chains.

Why nowStorage is the binding constraint on the transition; supply-chain sovereignty makes domestic chemistry strategic, not just commercial.

TailwindAntler (Meine Electric precedent), Deep Tech FoF, ACC-PLI.

ThesisA platform that finances, installs, and operates rooftop and community solar for households and SMEs that can't access capital today.

Why nowHardware costs have collapsed; the bottleneck is financing, trust, and O&M — a software-and-fintech problem.

TailwindAntler (Sustvest/Wiom precedent), state solar schemes.

ThesisAI and process engineering to cut energy and emissions in India's hard-to-abate heavy industry (steel, cement, chemicals).

Why nowCBAM-style border carbon rules threaten India's exports; efficiency is now a competitiveness issue, not a CSR line.

TailwindAccel (Haber precedent), RDI, export pressure.

ThesisIntelligence layer for India's two- and three-wheeler EV boom — charging networks, battery-health analytics, swapping, and second-life repurposing.

Why nowIndia electrifies from the bottom (2W/3W) up; the software and battery-data layer across this fragmented fleet is unbuilt.

TailwindFAME-successor schemes, mobility investors.

ThesisSensing, leak detection, treatment, and reuse infrastructure for the world's most water-stressed large economy.

Why nowWater scarcity is becoming the binding constraint on agriculture, industry, and cities simultaneously; almost nothing is instrumented.

TailwindJal Jeevan-adjacent funding, climate funds.

ThesisAI-and-robotics-enabled sorting, recycling, and waste-to-resource operations — turning India's waste crisis into feedstock.

Why nowWaste management is a named Accel action-model vertical and a massive, under-automated physical-economy problem.

TailwindAccel (waste-mgmt vertical), SBM, state programs.

ThesisScalable capture and utilization tuned to Indian industrial point-sources and cost realities.

Why nowIndia can't decarbonize heavy industry on efficiency alone; cost-appropriate capture is a long-cycle deeptech bet the state now funds.

TailwindRDI scheme, ANRF, Accel LeapTech.

ThesisA B2B cross-border commerce platform that fixes discovery, trust, and trade finance for India's fragmented ~$770B export base.

Why nowChina+1 is actively reshaping supply chains and the missing piece is an AI-first aggregator connecting Indian manufacturers to global buyers; this is an explicitly stated frontier idea.

TailwindAntler NEXT100 (Google partner thesis), export schemes.

ThesisA manufacturing OS — MES, quality, maintenance, scheduling — built AI-native for India's mid-market factories, most of which run on paper and spreadsheets.

Why nowChina+1 is pulling manufacturing into India faster than the software to run it efficiently exists; legacy MES optimizes for tonnage, not flexibility.

TailwindAccel (manufacturing action models), PLI schemes.

ThesisTightly integrated design-manufacturing-logistics that lets Indian hardware teams iterate physical parts in days, not weeks — closing the Shenzhen gap.

Why nowYC has made this one of its most specific talent-targeted requests; whoever builds the fast iteration loop unlocks every downstream hardware company.

TailwindYC RFS (hardware speed), Accel LeapTech.

ThesisAffordable vision-and-robotics retrofits that bring automated inspection and assembly to factories that can't afford German robots.

Why nowNewer model architectures widen the range of automatable factory tasks; precision manufacturing for auto/aerospace is a live Atoms cohort theme.

TailwindAccel Atoms (LevelPlane precedent), PLI, Deep Tech FoF.

ThesisAI agents that run indirect and direct procurement — sourcing, RFQ, supplier discovery, negotiation — for mid-market manufacturers, priced on outcomes.

Why nowProcurement is high-spend, low-software, and structurally outsourced — a textbook services-as-software wedge that compounds with the export and manufacturing build-out.

TailwindServices-as-software thesis, manufacturing growth.

ThesisThe AI layer that makes physical environments — factories, warehouses, ports, construction sites — legible and controllable through cameras, sensors, and 3D understanding.

Why nowAI has saturated the digital world while the physical world stays "dumb"; spatial intelligence is the next frontier and a named Antler thesis.

TailwindAntler NEXT100 (explicit), Accel (construction).

ThesisPlanning, monitoring, and risk software for India's trillion-dollar infrastructure pipeline — catching delays and overruns before they compound.

Why nowConstruction is a named Accel vertical and one of the least digitized, highest-waste sectors; the infra spend is real and growing.

TailwindAccel (construction action models), Gati Shakti.

ThesisFully indigenous UAVs and counter-UAS — surveillance, logistics, and loitering systems — designed and built in India for the armed forces and exports.

Why nowIndia has shifted decisively to indigenous drones (deployed operationally), with iDEX writing real contracts; this is a closed-to-open-market transition happening now.

TailwindiDEX / ADITI / DISC, defence PLI.

ThesisAutonomy stacks, targeting, and edge intelligence for unmanned ground, air, and naval systems that must work without connectivity.

Why nowModern conflict is software-defined; India is actively funding startups to fill the autonomy gap rather than import it.

TailwindiDEX, ADITI, ANRF.

ThesisGPS-independent navigation (quantum/inertial) for defence and critical infrastructure that can't depend on foreign satellite signals.

Why nowA startup already won a $3M ADITI challenge to build quantum positioning for the Navy; the strategic pull is explicit.

TailwindADITI, National Quantum Mission.

ThesisA constellation-and-analytics company turning satellite imagery into decisions for agriculture, insurance, defence, urban planning, and climate.

Why nowIndia liberalized geospatial data collection and opened space to private players; the analytics layer on top is wide open.

TailwindIN-SPACe, geospatial policy.

ThesisCheaper access to orbit and indigenous small-sat platforms riding the global reusable-launch capacity boom.

Why nowIN-SPACe authorizes private launch and manufacturing; global launch capacity is about to expand enormously, and India has the cost base.

TailwindIN-SPACe, Deep Tech FoF, specialist space funds.

ThesisRadiation-, mass-, and thermal-optimized inference hardware and edge processing for the coming wave of satellites.

Why nowReusable rockets are about to make space compute a real market; YC explicitly calls for inference chips in space — India can compete on the design layer.

TailwindYC RFS (chips in space), ISM, IN-SPACe.

ThesisAutonomous surface and undersea vehicles plus maritime-domain awareness for India's long coastline and blue-economy ambitions.

Why nowMaritime is strategically and commercially under-served; dual-use creates both a defence buyer and a commercial market.

TailwindiDEX (Navy), Accel (maritime).

ThesisIndia-designed silicon for smart vision, IoT, automotive, and edge-AI — the design-led path to semiconductor relevance without owning a fab.

Why nowThe India Semiconductor Mission and DLI scheme are funding exactly this; a startup in this space already raised ₹107 Cr of VC on the back of the program.

TailwindISM + DLI, Deep Tech FoF, RDI.

ThesisSilicon designed for how agents actually run — looping, branching, bursty tool-use — rather than single prompt-in/response-out inference.

Why nowYC flags that current GPUs hit only 30–40% utilization on agentic workloads; this is a global architectural opening India's design talent can attack.

TailwindYC RFS (agent chips), ISM.

ThesisAdvanced packaging, materials, and assembly-test for the chips India is starting to make — the less glamorous but unavoidable layer of the stack.

Why nowFabs alone don't make a semiconductor industry; packaging and materials are where India can realistically win first.

TailwindISM, Deep Tech FoF.

ThesisQuantum sensing, components, and the post-quantum cryptography migration that every bank and government will be forced into.

Why nowThe National Quantum Mission is funding the ecosystem; post-quantum security is a near-term, regulation-driven market.

TailwindNational Quantum Mission, RDI.

ThesisRobots and actuators built for India's cost base and tasks — warehouse, agriculture, cleaning, eldercare — where global humanoids are priced out.

Why nowThe ecosystem for hardware/robotics has matured to where this is finally buildable in India, a stated Antler conviction.

TailwindAntler (robotics), Accel LeapTech, Deep Tech FoF.

ThesisNew materials — batteries, composites, green cement/steel, rare-earth alternatives — discovered and scaled with AI and domestic feedstock.

Why nowMaterials are upstream of energy, defence, and manufacturing sovereignty; AI-accelerated discovery shortens the brutal timelines.

TailwindRDI, ANRF, BioE3 (bio-materials).

ThesisDomain copilots that accelerate electronics, materials, and mechanical engineering R&D — reading papers, running simulations, proposing designs.

Why nowAccel explicitly wants "AI research assistants to accelerate R&D"; India's large engineering research base is starved of modern tooling.

TailwindAccel Atoms (explicit), ANRF.

ThesisA personalized AI tutor that adapts to each child's level and language, equally usable in rural India and a metro — making expert tutoring a right, not a privilege.

Why nowThis is a stated defining-shift thesis; the model quality and vernacular voice layer now make it real.

TailwindAntler NEXT100 (universal-expertise), IndiaAI.

ThesisSkilling and placement infrastructure that moves millions into AI-adjacent and AI-resilient work as automation reshapes the job base.

Why nowYC names "retraining workers for the AI economy" as a frontier request; India's demographic dividend turns into a liability without it.

TailwindYC RFS, Skill India, state programs.

ThesisSimulation-, AR-, and voice-driven training for trades — electricians, machinists, healthcare aides, drivers — at the scale India's manufacturing build-out demands.

Why nowThe factories and infrastructure are coming faster than the skilled labor; vernacular, hands-on AI training closes the gap.

TailwindPLI-linked skilling, manufacturing growth.

ThesisPrograms that get paid when learners get jobs — bundling AI instruction, real projects, and employer pipelines.

Why nowThe degree-to-employment link is broken; outcome-aligned models win trust and unit economics.

TailwindEdtech investors, NEP reforms.

ThesisTools that multiply the average teacher — lesson planning, grading, multilingual content, individualized attention for 50-child classrooms.

Why nowYou can't hire your way to quality at India's scale; augmenting teachers is the only lever that reaches every child.

TailwindIndiaAI, state ed-departments, Peak XV (Anandi precedent).

ThesisA control tower that optimizes routing, fleet, and middle-mile across India's notoriously fragmented, multi-modal, mostly-offline logistics.

Why nowLogistics is a named Accel action-model vertical; the inefficiency is enormous and the data to fix it is finally being captured.

TailwindAccel (logistics), Gati Shakti, mobility funds.

ThesisAI-powered safety, telematics, and driver-assist hardware for India's two-wheelers, trucks, and commercial fleets.

Why nowIndia has the world's deadliest roads and a giant fleet; affordable AI mobility hardware is a live Antler bet (Cautio precedent).

TailwindAntler (mobility hardware), road-safety policy.

ThesisFlying taxis and short-haul air mobility for India's congested megacities and difficult terrain.

Why nowA "LeapTech" bet with long timelines that the ecosystem is now willing to underwrite (Accel-backed precedent already flying prototypes).

TailwindAccel Atoms X / LeapTech, IN-SPACe-adjacent regulation.

ThesisAffordable automation for the quick-commerce and e-commerce fulfillment build-out — picking, sorting, inventory at Indian cost points.

Why nowQuick commerce exploded; the labor-heavy fulfillment layer is the next thing to be automated.

TailwindRobotics funds, consumer-internet tailwind.

ThesisSoftware and micro-logistics that make it economic to deliver goods, medicine, and services to India's villages.

Why nowONDC and rural consumption growth open the demand; the cost-to-serve is the unsolved problem.

TailwindONDC, rural-commerce investors.

ThesisSoftware to run India's ports and shipping — scheduling, customs, container flow — as trade volumes scale.

Why nowPorts are bottlenecks for the export thesis; maritime is a named, under-built Accel vertical.

TailwindAccel (maritime), Sagarmala.

ThesisTools that process the deluge of government forms, applications, and citizen queries — replicating Estonia-style digital government at India's scale.

Why nowYC explicitly calls for AI for government; the customers are sticky and the contracts expand, and India's DPI gives a uniquely strong base to build on.

TailwindYC RFS (AI for government), Digital India.

ThesisAn AI assistant that tracks, updates, retrieves, and submits a citizen's documents — making banking, taxes, property, and registrations as frictionless as UPI made payments.

Why nowUPI and DigiYatra proved the model for one domain each; the general-purpose document layer is a named NEXT100 thesis and still unbuilt.

TailwindAntler NEXT100 (explicit), DPI rails.

ThesisProtocols and infrastructure that extend the India Stack into new domains — open networks for logistics, health records, or skilling, the way ONDC did for commerce.

Why nowIndia's DPI playbook is proven and being exported; the next rails are greenfield and have explicit state backing.

TailwindDPI ecosystem, MeitY, ONDC.

ThesisTools that help clear India's tens-of-millions-strong case backlog — translation, summarization, scheduling, legal research for judges and litigants.

Why nowThe backlog is a national crisis; AI translation and summarization across Indian languages can compress timelines dramatically.

TailwindeCourts, IndiaAI, govtech procurement.

ThesisReal-time surveillance, outbreak prediction, and supply-chain tooling for India's public-health system.

Why nowPost-pandemic political will plus AI plus geospatial data make population-scale early warning achievable.

TailwindAyushman Bharat Digital Mission, ICMR, BIRAC.

ThesisSoftware to run cities — traffic, utilities, waste, safety — integrating the sensors and data that smart-city programs deployed but never connected.

Why nowIndia is urbanizing at unprecedented speed; the hardware is partly there, the intelligence layer isn't.

TailwindSmart Cities legacy, state programs.

ThesisA dedicated AI representative for every brand and product — onboarding, supporting, and personalizing the entire post-purchase lifecycle, restoring the human touch to commerce at scale.

Why nowE-commerce traded experience for scale; agentic AI lets you have both, a stated NEXT100 thesis.

TailwindAntler NEXT100 (explicit), ONDC.

ThesisGeneration and distribution tools for India's creator economy in 20+ languages — video, audio, dubbing, localization.

Why nowIndia's content consumption is exploding in regional languages while tooling remains English-first; AI dubbing/generation unlocks the long tail.

TailwindIndiaAI, consumer-internet funds.

ThesisAffordable AI-native devices — health wearables, companions, learning devices — designed for Indian price points and use cases.

Why nowAntler explicitly flags consumer AI hardware (Ultrahuman-class numbers) as buildable in India now; the supply chain and demand have arrived.

TailwindAntler (consumer hardware), ISM.

ThesisReputation, authenticity, and discovery infrastructure for the millions of small sellers entering open commerce — fighting counterfeits and building buyer trust.

Why nowONDC unbundles the trust that platforms used to provide; someone has to rebuild it as shared infrastructure.

TailwindONDC, consumer-fintech crossover.

ThesisThe identity and policy layer for a world where autonomous agents act on behalf of people and companies — who is this agent, what may it do, who is liable.

Why nowAgentic systems are being deployed before the governance to control them exists; this is foundational infrastructure, not a feature.

TailwindEnterprise + DPI security demand, Peak XV (auth precedent).

ThesisMonitoring, tracing, and assurance tooling that tells enterprises and regulators what their AI systems actually did and why.

Why nowRegulated Indian enterprises and the IndiaAI responsible-AI framework will mandate explainability; the tooling is nascent.

TailwindIndiaAI (responsible AI), enterprise compliance.

ThesisInfrastructure that operationalizes India's Digital Personal Data Protection Act — consent management, data minimization, cross-border compliance — for every company that touches personal data.

Why nowThe DPDP Act is now live and enforceable; compliance is mandatory, complex, and currently unserved by India-specific tooling.

TailwindDPDP Act, Account Aggregator, enterprise demand.