Two Bets on India’s Financial and Medical Infrastructure: Dognosis and Rezolv

Two announcements from India’s startup ecosystem this week sit at opposite ends of the technology spectrum — one involves beagles on a farm outside Bengaluru, the other involves AI agents conducting 6.5 million minutes of borrower conversations every month. What connects them is a shared ambition: to use artificial intelligence to solve structural problems in sectors that have long suffered from inadequate infrastructure. Dognosis is trying to make cancer screening accessible to the hundreds of millions of Indians who currently have no access to it. Rezolv is trying to make India’s lending system work better for the 22 banks and NBFCs that have already partnered with it. Both raised capital this week. Both are worth understanding in detail.

Dognosis: teaching AI to read what dogs already know

On a two-acre farm on the outskirts of Bengaluru, a team of beagles, labradors, and Dutch shepherds report for work every day. Their job is to detect cancer — and according to the clinical data available so far, they are surprisingly effective at it.

Dognosis, a two-year-old Bengaluru-based startup, is combining trained detection dogs with AI-powered sensor arrays to identify early signs of cancer from a single human breath sample. The product is called BreathEasy. The process is straightforward at the patient’s end: a person breathes into a face mask for approximately ten minutes. The mask captures volatile organic compounds — chemical signals present in exhaled breath that shift measurably in the presence of diseases including cancer. The sealed mask is then couriered to Dognosis’s facility, where the trained dogs sniff each sample in a controlled laboratory environment. As they do, sensors and AI systems read the animals’ movements, respiration, body language, and physiological responses — translating a pause, an alert posture, or a change in breathing pattern into structured diagnostic data.

The scientific foundation for this approach is well-established. More than 40 double-blind trials published in peer-reviewed journals have demonstrated that dogs can detect various diseases — including multiple types of cancer — with accuracy that often exceeds conventional screening methods. Dogs’ olfactory systems contain approximately 300 million scent receptors, compared to around 6 million in humans, and they can detect chemical concentrations in the parts-per-trillion range. The question Dognosis is answering is not whether dogs can detect cancer — that is settled — but whether the detection can be standardised, scaled, and made reliable enough to function as a mass screening tool.

The co-founder Kulgod has put the challenge precisely: the bottleneck was never whether dogs could smell cancer. It was how to standardise and scale the dog’s nose. The AI layer is how Dognosis is solving that bottleneck. Rather than relying on a human handler’s interpretation of a dog’s behaviour — an approach that introduces inconsistency — the system uses sensors to capture objective physiological data from the animal and AI to translate that data into a consistent output. The result, according to Phase 2 trial data published in the Journal of Clinical Oncology, is 90% sensitivity and specificity — meaning that if someone has cancer, the dogs identify it correctly approximately 90% of the time, and the false positive rate is commensurately low.

“There is a 90 per cent sensitivity and specificity, which means that if someone has cancer, the dogs are able to identify it around 90 per cent of the time.” — Dr Swaratika Majumdar, Dognosis

The Phase 2 trial covered 1,500 participants. In April 2026, Dognosis kicked off its Phase 3 trial — a 10,000-person study across 10 hospitals in India over 12 months. The trial will test asymptomatic people at higher risk of cancer, and survivors at risk of recurrence. Participants flagged by the screening will receive conventional follow-up including imaging and biopsy. If the Phase 3 results hold, Dognosis plans to launch commercially in Bengaluru in 2027 before expanding nationally through partnerships with diagnostic chains and healthcare providers.

One aspect of the regulatory situation deserves specific attention. Kulgod has noted that the test does not require Indian regulatory approval as a prescreening tool — only diagnostic devices are regulated in India, and BreathEasy is positioned as a prescreening tool that flags individuals for further conventional testing, rather than providing a clinical diagnosis. This is both a regulatory advantage and an important caveat: BreathEasy, if it works at scale, tells a patient that they should see an oncologist. It does not tell them they have cancer. That distinction matters for how the product is communicated to users and healthcare partners.

The masks can retain potency for up to three months after collection, and potentially longer — a logistics advantage that means samples can be shipped hundreds or thousands of kilometres before being analysed, making centralised processing at a single facility viable even for a national screening programme.

Dognosis is not the only company globally pursuing this model. SpotitEarly, a US-based biotech that presented at TechCrunch Disrupt 2025, is building an at-home cancer test using the same dogs-plus-AI approach for Western markets. The existence of a well-funded US counterpart validates the thesis globally. Dognosis’s India-specific advantage is the scale of the unaddressed screening gap it is targeting — and the cost structure that makes low-cost mass screening commercially viable in India in a way it may not be in higher-cost markets.

The cancer screening gap Dognosis is building into

India’s cancer burden is significant and growing. The country accounts for approximately 10% of all cancer cases globally, with an estimated 1.46 million new cancer cases and 900,000 deaths annually. The five-year survival rate for most cancers in India is substantially lower than in developed economies — a gap driven almost entirely by late-stage diagnosis. More than 60% of cancer cases in India are detected at Stage 3 or Stage 4, when treatment options are limited and outcomes are poor. If those same cancers were detected at Stage 1 or Stage 2, five-year survival rates for most cancer types would exceed 80 to 90%.

The conventional cancer screening tools available in India — colonoscopy, mammography, CT scan, biopsy — are expensive, invasive, and concentrated in major urban centres. A colonoscopy costs ₹5,000 to ₹15,000. A CT scan costs ₹3,000 to ₹8,000. For the 65% of Indians who live outside major cities, accessing these tests requires travel, expense, and time that most people cannot or will not commit to for a screening test when they feel healthy.

Dognosis’s BreathEasy test, if commercially launched at the price points the company has indicated it is targeting, would represent a fundamentally different cost and accessibility profile — a non-invasive breath test that can be administered anywhere, with a sample couriered to a centralised facility. The commercial model and pricing have not yet been publicly confirmed, pending Phase 3 results. But the structural opportunity is clear: a low-cost, non-invasive cancer prescreening tool with 90% sensitivity, available nationally through diagnostic chains, would address a gap in India’s health infrastructure that no existing product or service currently fills.

Rezolv: building the AI operating system for India’s lenders

While Dognosis is solving a problem in healthcare infrastructure, Rezolv is solving one in financial infrastructure — and doing so at a pace that has attracted Norwest Venture Partners, one of the most established US venture firms active in India, to lead its Series A.

Rezolv has raised $12.5 million in a Series A round led by Norwest, with participation from Vertex Ventures Southeast Asia and India, and existing investor 3one4 Capital. The round values the company at $51.2 million — a roughly fourfold increase from its seed round valuation of $12.8 million, which was set just over a year ago when 3one4 Capital led a $3.5 million seed investment. Total funding raised stands at $16 million.

The founding team is immediately credible in the domain. Karan Mehta and Sonali Jindal previously co-founded Kissht, a digital consumer lending platform that operates at scale in India. They bring not just technology understanding to Rezolv, but direct experience of what India’s lending infrastructure looks like from the lender’s side — what breaks, what scales, and where AI can create measurable commercial value rather than incremental operational improvement.

Rezolv launched in October 2024 and reached an annualised revenue run rate of approximately ₹30 crore ($3.5 million) by March 2026 — the end of its first full year of operations. For a B2B enterprise software company serving regulated financial institutions, that is a strong commercial velocity: 22 lenders signed, including ICICI Bank and AU Small Finance Bank, in under 18 months of operations.

The platform covers the full lending lifecycle — sales, risk assessment, underwriting, collections, recoveries, and field operations — through AI-driven automation. Its Strategy Builder product has delivered a 35% improvement in bounce and resolution rates for collections. The company conducts 6.5 million minutes of AI-powered borrower conversations every month and manages collections across more than 12 million loan accounts. These are not aspirational figures — they represent live deployments in production at regulated financial institutions, which have compliance and operational requirements that make it considerably harder to deploy new technology than in most enterprise software categories.

“AI adoption is no longer the challenge as nearly every organisation today is implementing AI. The real challenge is metricisation: can you quantify the business impact AI is creating?” — Sonali Jindal, Co-founder, Rezolv

The framing from Jindal is precise and worth taking seriously. The AI implementation conversation in Indian banking has moved past the stage where institutions are asking whether to adopt AI. The question now is whether the AI they are adopting is generating measurable outcomes — in recovery rates, in cost per collection, in resolution speed, in workforce productivity. Rezolv’s entire commercial proposition is built around answering that question with specific, auditable numbers rather than directional claims. The 35% improvement in bounce and resolution rates is the kind of metric that a CFO or CRO at a bank can take to their board. That is what differentiates a platform that gets deployed in production from one that stays in a pilot.

The investors — and what their backing signals

Norwest Venture Partners, which led the Series A, is a $15 billion AUM multi-stage investment firm with a significant India portfolio. Their decision to lead a Series A in a Mumbai-based B2B fintech company founded less than two years ago — at a $51.2 million valuation — reflects high conviction in both the founding team’s track record at Kissht and the commercial traction Rezolv has demonstrated in its first full year. Norwest has historically backed companies that are building infrastructure-layer products in large, regulated markets — Rezolv fits that profile precisely.

Vertex Ventures Southeast Asia and India brings regional portfolio context — the firm has backed several fintech and B2B software companies across India and Southeast Asia, and its participation signals confidence in Rezolv’s potential to expand beyond India into regional markets, a stated objective for the use of the Series A capital.

3one4 Capital, which led the seed and participated in the Series A, has deepened its commitment. Akash Sharma, Principal at 3one4, described the round as validating a thesis the firm backed with conviction — the shift from services to product as AI transforms debt collection, and Rezolv’s full-lifecycle platform capturing that shift. The double-down from a seed investor at Series A, when they have visibility into actual operating data from the past 18 months, is a stronger signal than the initial seed check.

Rezolv’s primary competitors are Credgenics and DPDzero, both of which have raised capital for AI-powered debt collection platforms targeting the same bank and NBFC customer base. DPDzero raised $7 million in 2025 to develop multilingual AI agents for collections. The competitive dynamic is worth watching: Rezolv’s differentiation is in the breadth of its platform — covering sales, underwriting, and risk alongside collections — rather than depth in collections alone. The question the Series A capital will answer is whether that broader platform strategy translates into larger contract sizes and higher switching costs, or whether lenders prefer point solutions from specialists in each workflow.

What both companies have in common

Dognosis and Rezolv are building in very different domains. But the structural logic underpinning both of them is the same: India has large, well-defined problems in healthcare and financial services that have been inadequately addressed by incumbent solutions, and AI — applied correctly, with domain expertise, at the right price point — can create genuinely new value rather than incremental improvement.

Dognosis is not trying to build a better version of an existing cancer screening tool. It is trying to create a screening capability that does not currently exist at scale for the vast majority of Indians — non-invasive, low-cost, accessible outside major cities, applicable to multiple cancer types from a single test. Rezolv is not trying to build a better call centre software. It is trying to replace the entire workflow stack that Indian lenders currently use to manage their loan portfolios — from origination through collections — with an AI-native platform that delivers measurable outcomes rather than productivity claims.

Both are early. Dognosis has Phase 3 trial results to generate before it can launch commercially. Rezolv has ₹30 crore in ARR and needs to scale to the revenue level that justifies a $51.2 million valuation. But the direction of travel in both cases is clear — and the investors backing them have made a specific bet that the infrastructure gaps they are targeting are real, large, and now addressable in ways they were not five years ago.

Sources: Business Standard, The Next Web, Dog Express, New Kerala, Taipei Times, YourStory, Inc42, Dealroom, Startup Talky, Deal Street Asia, TN Global, Viestories. Data verified as of August 18, 2026.