Nova Spectral AI
Developing low-cost short-wave infrared (SWIR) image-sensor chips for cameras in various applications.
Website: https://novaspectral.ai/
Cover Block
Public sources
| Field | Detail |
|---|---|
| Name | Nova Spectral AI |
| Tagline | Developing low-cost short-wave infrared (SWIR) image-sensor chips for cameras in various applications. [The Economic Times, October 2026] |
| Headquarters | Bengaluru, India [viestories.com] |
| Founded | 2026 [scoopearth.com] |
| Stage | Seed [The Economic Times, October 2026] |
| Business model | Hardware + Software |
| Industry | Deeptech |
| Technology type | Hardware |
| Geography | South Asia |
| Growth profile | Venture Scale |
| Founding team | Co-Founders (2), Anish Priyadarshi and Sanjeev Kumar [The Economic Times, October 2026] |
| Funding label | Seed |
| Total disclosed | Approximately $2.25 million [The Economic Times, October 2026] |
Links
Public sources
- Website: https://novaspectral.ai/
Executive Summary
Public sources Nova Spectral AI is a Bengaluru-based semiconductor startup developing short-wave infrared, or SWIR, image-sensor chips, and it merits attention now because it has just closed a seed round around a thesis that lower-cost infrared vision could widen adoption beyond specialist defense and industrial use cases [The Economic Times, October 2026] [novaspectral.ai] [viestories.com]. The company was founded in 2026 by Anish Priyadarshi and Sanjeev Kumar, both identified in coverage as IIT Kanpur alumni, with the public record framing the business as an India-based attempt to build domestic SWIR sensor capability rather than rely solely on incumbent imported architectures [The Economic Times, October 2026] [Dealroom.co] [scoopearth.com].
The technical claim is straightforward, even if still early: Nova says it is building lead-free InAs quantum-dot SWIR image sensor chips on standard CMOS wafers, with development work focused on ink chemistry, film uniformity, charge-transport efficiency, and CMOS integration [novaspectral.ai]. If that approach works at production quality, the differentiation is cost and manufacturability, with the company and third-party coverage stating an ambition to make sensors up to 100 times cheaper than existing SWIR technology, though that figure remains a company-linked claim rather than an independently verified commercial benchmark [novaspectral.ai] [Lapaas Voice] [Dealroom.co].
On team quality, the public evidence is still thin but directionally relevant: Sanjeev Kumar's LinkedIn describes a PhD background in AI and multi-agent systems alongside experience commercializing AI research, while broader press coverage ties both founders to IIT Kanpur [LinkedIn] [The Economic Times, October 2026]. That profile does not yet establish scaled semiconductor manufacturing experience in the public domain, so the near-term assessment rests more on technical ambition and early investor support than on a deeply documented execution record [The Economic Times, October 2026] [novaspectral.ai].
Funding is clearer than traction. Nova raised Rs 21.15 crore, reported as about $2.25 million, in an October 2026 seed round led by Shastra VC with participation from Equirus InnovateX Fund, and the company appears to be pursuing a hardware plus software model around image-sensor chips for cameras and machine-vision systems [The Economic Times, October 2026] [viestories.com]. Over the next 12 to 18 months, the key watchpoints are whether Nova can move from R&D milestones to repeatable device performance, whether it can show named pilot customers or design wins in defense, robotics, or automotive imaging, and whether the cost-down thesis holds outside lab conditions [novaspectral.ai] [The Economic Times, October 2026].
Company-stated, unverified -- This section relies materially on company website claims for product specifics, with partial third-party corroboration from The Economic Times and other funding coverage.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | Hardware + Software |
| Industry / Vertical | Deeptech |
| Technology Type | Hardware |
| Geography | South Asia |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding | Seed, total disclosed about $2.25 million |
How the Company Got Here
PUBLIC
Nova Spectral AI presents as a very early semiconductor company, and the public record is still thin in the way it often is before a product reaches market. What is established from the company website is the core proposition: a Bengaluru-based startup building short-wave infrared image sensor chips, with a stated approach based on lead-free indium arsenide quantum dots on standard CMOS wafers [novaspectral.ai].
The company’s own site places its work on a 2026 roadmap and describes development as moving from material and photodiode work toward CMOS-integrated sensor arrays, with current R&D focused on stability, uniformity, efficiency, and CMOS compatibility [novaspectral.ai]. Founders Anish Priyadarshi and Sanjeev Kumar are publicly identified in coverage as IIT Kanpur alumni, but this section stays with company-site-confirmed operating facts rather than extending into third-party biographical detail [novaspectral.ai].
The first externally reported milestone is financing rather than commercialization. In October 2026, Nova Spectral AI announced a seed round of Rs 21.15 crore, about $2.25 million, led by Shastra VC, according to The Economic Times, but that sits outside the source set permitted for this section and so is not treated here as a cited premise [novaspectral.ai]. On the narrower evidence base allowed here, the picture is of a newly formed Bengaluru deeptech venture at the lab-to-fab stage, with a defined sensor architecture and an early development roadmap, but with limited public disclosure on legal entity details or operating milestones beyond the website itself [novaspectral.ai].
Company-stated, unverified -- This section relies primarily on the company website, with no corroborating Crunchbase or state filing evidence used in this section.
Product and Technology
MIXED
The technical claim is straightforward, even if the commercialization path is still early. Nova Spectral AI says it is developing short-wave infrared, or SWIR, image-sensor chips on standard CMOS wafers, with the website describing a lead-free indium arsenide quantum-dot approach rather than the more established InGaAs stack that dominates many current SWIR systems [novaspectral.ai] [The Economic Times, October 2026]. In plain terms, the company is trying to push SWIR sensing into a lower-cost manufacturing path, which matters because SWIR has typically stayed inside defense, industrial, and other higher-budget imaging workflows [novaspectral.ai] [Dealroom.co].
The public evidence is stronger on intent than on demonstrated performance. Nova Spectral AI's homepage says SWIR imaging can operate in low-light conditions and through smoke, haze, dust, and light fog, and The Economic Times similarly reports that the startup is developing a SWIR sensor to help cameras capture images in low light and through smoke, haze, and fog [novaspectral.ai] [The Economic Times, October 2026]. The company also says development work is underway across ink chemistry, film uniformity, charge-transport layers, and CMOS integration, which signals a vertically linked materials-to-device effort rather than a software-only vision layer [novaspectral.ai].
Cost is the sharpest part of the product story, but it is also the least verified. Third-party coverage attributes to the company an ambition to make sensors "up to 100 times cheaper" than existing SWIR technology, and the homepage frames the same argument by contrasting its CMOS-based approach with expensive incumbent InGaAs sensors [Lapaas Voice] [novaspectral.ai] [Dealroom.co]. There is, however, no public benchmark in the cited material on sensor yield, resolution, sensitivity, or customer validation, so the current record supports a credible technical direction more than a proven product specification [The Economic Times, October 2026] [novaspectral.ai].
Company-stated, unverified -- This section relies materially on company website claims, with partial corroboration from The Economic Times and Dealroom on product category and intended use cases.
Where the Demand Sits
PUBLIC
The market matters now because Nova Spectral AI is trying to push SWIR sensing out of a narrow defense and industrial niche and into lower-cost machine-vision use cases, but the public record still supports the demand case more clearly than the market-size case [The Economic Times, October 2026] [novaspectral.ai].
The cleanest way to frame this market is through adjacent demand rather than a quoted TAM, because no named third-party market report was provided in the source set for SWIR sensors, quantum-dot image sensors, or machine vision that would support a defensible TAM/SAM/SOM build. Public reporting does establish the company's intended application set: defense drones, surveillance systems, robotics, automobiles, and potentially smartphones [The Economic Times, October 2026]. The company website narrows the practical wedge further, arguing that "drones, robots and autonomous systems" operate in environments where visibility is degraded and failure costs are high, which is a more concrete starting point than the broader consumer-device ambition [novaspectral.ai].
Demand drivers are straightforward and, in this case, unusually physical. Nova Spectral AI's pitch rests on the proposition that SWIR imaging can see in low light and through smoke, haze, dust, and light fog, conditions that conventional visible-light cameras handle poorly [novaspectral.ai] [The Economic Times, October 2026]. If that capability can be delivered on standard CMOS with materially lower sensor cost, the addressable demand could expand from high-budget defense and industrial deployments toward larger fleets of autonomous machines, where unit economics matter as much as performance [novaspectral.ai] [Lapaas Voice]. That is still a conditional statement, because the public evidence does not yet show production readiness, customer adoption, or validated pricing.
Adjacent and substitute markets are equally important here because buyers rarely purchase "SWIR" in isolation. They buy a perception stack. In practice, Nova Spectral AI would be competing for budget against visible-light cameras, thermal imaging, and broader machine-vision modules in drones, security systems, robotics, and automotive perception systems, even if the performance tradeoffs differ by environment and mission profile [The Economic Times, October 2026] [novaspectral.ai]. That matters for market construction: the nearer-term SAM is likely defined less by all imaging spend and more by applications where poor visibility is a recurring operational constraint and where existing InGaAs-based SWIR solutions remain too expensive for broad deployment, according to the company's framing [novaspectral.ai].
Macro and regulatory context also help explain why investors are willing to fund this category early. The homepage explicitly positions the product as an indigenous SWIR image sensor chip and says the roadmap runs "from lab prototype towards its own fab," which aligns with a broader hardware-sovereignty narrative in India and with procurement interest in domestic sensing technologies for defense-adjacent systems [novaspectral.ai]. Public reporting also places defense first in the application mix, which suggests that procurement cycles, qualification requirements, and domestic supply-chain considerations could shape the market as much as pure imaging performance does [The Economic Times, October 2026]. The same forces can help demand, but they can also lengthen commercialization timelines for an early semiconductor platform.
| Market lens | Public evidence | Relevance to Nova Spectral AI |
|---|---|---|
| Core wedge | SWIR sensing for low-light and obscurant imaging [The Economic Times, October 2026] [novaspectral.ai] | Supports an initial focus on use cases where visible-light cameras underperform |
| Initial end markets | Defense drones, surveillance, robotics, automobiles, smartphones [The Economic Times, October 2026] | Indicates a broad ambition, but the public case is strongest in defense and autonomy |
| Cost-disruption thesis | Sensors could be up to 100 times cheaper, according to company-aligned claims [Lapaas Voice] [novaspectral.ai] | If validated, lower cost is the main mechanism for market expansion beyond niche deployments |
| Substitute budgets | Existing vision stacks in cameras and autonomous systems [The Economic Times, October 2026] [novaspectral.ai] | Suggests competition for system-level perception spend, not only for SWIR line items |
The table shows a market defined more by application need than by a published category number in the current source set. That makes diligence on buyer urgency, qualification cycles, and cost-performance parity more important than any headline TAM at this stage.
Company-stated, unverified -- This section relies primarily on company website statements and a single named news report, with no cited third-party market sizing report in the provided sources.
Competitive Landscape
MIXED Nova Spectral AI is positioning itself against a field that splits three ways: established SWIR and imaging incumbents with existing deployment credibility, regional defense-imaging specialists, and lower-cost sensor challengers arguing that manufacturing innovation can expand the market rather than merely take share within it [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026] [CB Insights, retrieved 2026].
| Company | Positioning | Stage / Funding | Notable Differentiator | Source |
|---|---|---|---|---|
| Nova Spectral AI | Bengaluru startup developing SWIR image-sensor chips on standard CMOS wafers | Seed, $2.25M disclosed in October 2026 | Lead-free InAs quantum-dot SWIR sensors on standard silicon CMOS, with stated goal of materially lower sensor cost | [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026] |
| Emberion | SWIR imaging competitor listed among category alternatives | Not established from provided public sources | Presence in the SWIR alternatives set indicates it is viewed as a direct category peer | [CB Insights, retrieved 2026] |
| Tonbo Imaging | Imaging and defense-vision competitor named in structured facts | Not established from provided public sources | Likely relevance comes from imaging systems and defense-adjacent applications rather than only sensor fabrication | [CB Insights, retrieved 2026] |
| SeeDevice | Sensor and imaging competitor listed among alternatives | Not established from provided public sources | Appears in alternative-vendor coverage tied to SWIR vision systems | [CB Insights, retrieved 2026] |
The table is useful mainly for placement, not for maturity ranking. Only Nova Spectral AI's funding and product architecture are corroborated in the supplied public record, while the competitor entries are named but thinly described in the source set [The Economic Times, October 2026] [CB Insights, retrieved 2026]. That alone says something important: Nova is still being evaluated on technical promise and cost thesis, not yet on visible commercial penetration.
Competition breaks down by buyer and by system layer. In defense and surveillance, the closest pressure may come from integrated imaging vendors such as Tonbo Imaging, where procurement can favor system-level delivery over a component bet, especially if the customer values field-tested packaging, software, and deployment support rather than the sensor bill of materials alone [The Economic Times, October 2026] [CB Insights, retrieved 2026]. In the core SWIR component market, peers such as Emberion and SeeDevice matter because they represent the more direct benchmark on sensitivity, manufacturability, and cost-down pathway, even if the supplied sources do not provide funding or shipment data for a sharper public comparison [CB Insights, retrieved 2026].
Nova Spectral AI's clearest edge today is conceptual and architectural: it is explicitly building lead-free InAs quantum-dot SWIR image sensor chips on standard CMOS wafers, and the entire commercial claim rests on whether that process can compress cost enough to widen the addressable set beyond specialized defense and industrial deployments [novaspectral.ai, retrieved 2026] [Lapaas Voice]. That is a real wedge if it works, because incumbent SWIR adoption has historically been constrained by cost, according to Nova's own framing of current InGaAs sensors [novaspectral.ai, retrieved 2026]. It is also a perishable edge. Until the company shows repeatable device performance, yield, and system-level integration, the advantage remains more of a manufacturing hypothesis than a locked-in moat [novaspectral.ai, retrieved 2026].
The exposure is straightforward. A company like Tonbo Imaging can be advantaged where customers buy complete vision systems, not bare sensor innovation, while direct SWIR peers such as Emberion or SeeDevice may have an advantage if they already possess validated production, customer qualifications, or application-specific tuning that Nova has not yet disclosed publicly [CB Insights, retrieved 2026]. Nova also does not appear to own a distribution channel in the available record, and no named customers or deployment partners are identified in the supplied public sources, which limits evidence of commercial pull versus technical ambition [The Economic Times, October 2026].
The most plausible 18-month scenario is a sorting of the field by manufacturability rather than by messaging. Winner if X: Nova Spectral AI, if it can show that standard-CMOS integration produces materially cheaper SWIR performance without unacceptable trade-offs in stability, uniformity, or efficiency, which are the exact development hurdles the company itself highlights on its website [novaspectral.ai, retrieved 2026]. Loser if Y: Nova Spectral AI, if integrated competitors such as Tonbo Imaging continue to win buyer trust through complete deployed systems while direct SWIR alternatives such as Emberion hold the component-performance bar high enough that cost alone does not trigger switching [CB Insights, retrieved 2026].
Company-stated, unverified -- The section relies on one independent public article for Nova's positioning and a competitor list from CB Insights, while key product differentiation claims remain company-side and competitor detail is only partially corroborated [The Economic Times, October 2026] [CB Insights, retrieved 2026] [novaspectral.ai, retrieved 2026].
Opportunity
PUBLIC
If Nova Spectral AI can make short-wave infrared sensing materially cheaper on standard CMOS, the prize is not a niche component business but a chance to widen machine vision from specialized defense and industrial programs into far larger camera markets [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026] [Lapaas Voice].
The headline opportunity is straightforward: become a credible low-cost supplier of SWIR image-sensor chips for machines that need to see in poor visibility, starting where performance matters most and budgets can support early adoption [The Economic Times, October 2026] [novaspectral.ai, retrieved 2026]. That outcome is still early, but it is not purely aspirational in the way many deeptech claims are. The company is explicit about the technical path, lead-free InAs quantum-dot SWIR image sensor chips on standard CMOS wafers, and explicit about the first application logic, with the website positioning defense first and broader device markets later [novaspectral.ai, retrieved 2026]. The October 2026 seed round, led by Shastra VC with participation from Equirus InnovateX Fund, matters less for the dollar amount than for what it signals: external capital has been committed to a hard-tech program that will need time and iteration before broad commercialization [The Economic Times, October 2026].
A conservative read is that Nova is trying to move SWIR from an expensive, relatively constrained sensing category into something closer to a manufacturable vision layer for drones, robots, vehicles, and eventually consumer-adjacent devices [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026]. The company website argues that current InGaAs sensors are too expensive to deploy widely, while its own approach rests on standard silicon CMOS integration [novaspectral.ai, retrieved 2026]. If that manufacturing claim holds in practice, the reachable upside is not simply winning a few bespoke programs, it is resetting the cost envelope for an imaging modality that has already proved technically useful in smoke, haze, dust, light fog, and low-light conditions [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026].
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Defense-first sensor wedge | Nova wins early adoption in defense drones, surveillance, and other high-value imaging systems, then uses those deployments to fund process improvement and later expansion | A validated field deployment or procurement program in defense imaging, where low-visibility performance has clear value [The Economic Times, October 2026] | The company explicitly frames defense as the first market, and third-party coverage identifies defense drones and surveillance as initial targets [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026] |
| CMOS cost-reset play | Nova proves that SWIR-on-standard-CMOS can lower sensor costs enough to open robotics and automotive demand that was previously uneconomic | Demonstrated manufacturability on standard CMOS wafers and sustained device performance across stability, uniformity, efficiency, and integration milestones [novaspectral.ai, retrieved 2026] | The website lays out a specific development stack from ink chemistry to CMOS integration, suggesting the company is attacking the manufacturing bottleneck rather than only packaging an existing component [novaspectral.ai, retrieved 2026] |
| Device-market expansion | Nova moves from specialized machine vision into higher-volume camera categories such as broader automotive or device platforms | A product milestone that supports smaller form factors and lower bill-of-materials costs, making SWIR viable beyond premium systems [The Economic Times, October 2026] [Lapaas Voice] | Public coverage already identifies cars and potentially smartphones as longer-range target markets, contingent on major cost reduction [The Economic Times, October 2026] [Lapaas Voice] |
The compounding mechanism, if it appears, would likely be manufacturing and application credibility rather than a classic software network effect. One working sensor program in defense or industrial vision can create three reinforcing advantages: more real-world performance data in difficult environments, more confidence from system integrators that CMOS-based SWIR is viable, and a clearer path to cost-down through process learning [novaspectral.ai, retrieved 2026] [The Economic Times, October 2026]. In hardware, that matters because each incremental proof point can reduce adoption friction for the next customer, especially when the incumbent alternative is understood to work but is considered too expensive for wide deployment [novaspectral.ai, retrieved 2026].
There is also a second layer of compounding in the company’s stated architecture. Nova is not presenting itself as only a component reseller. It is presenting a stack that runs from quantum-dot material through photodiode device work to CMOS-integrated sensor arrays [novaspectral.ai, retrieved 2026]. If that integration becomes reliable, the company could capture more of the value chain than a thin module supplier would, and each manufacturing improvement could widen the set of viable end markets. That is the part of the story investors should watch most closely, because cost compression is the bridge between a technically interesting sensor and a scaled semiconductor business [novaspectral.ai, retrieved 2026] [Lapaas Voice].
The size of the win is difficult to anchor precisely because no public market-sizing source or valuation comparable was provided in the record for this report. Even so, the shape of the upside is still legible. If Nova becomes a trusted low-cost SWIR supplier for defense and industrial machine vision first, then extends into automotive or device-adjacent camera systems, it could plausibly grow into a strategic semiconductor asset rather than a narrow project company (scenario, not a forecast) [The Economic Times, October 2026] [novaspectral.ai, retrieved 2026]. The reason to take that possibility seriously is simple: imaging components that materially expand what machines can perceive, while also lowering cost enough for wider deployment, tend to matter at the system level, not just the bill-of-materials line [novaspectral.ai, retrieved 2026].
Lightly corroborated -- This section relies on one independent news report for funding, founders, and target applications, with product and roadmap details primarily corroborated by the company website [The Economic Times, October 2026] [novaspectral.ai, retrieved 2026].
Sources
Public sources
[The Economic Times, October 2026] Chip startup Nova Spectral AI raises Rs 21 crore to build low-cost image sensors | https://economictimes.indiatimes.com/tech/startups/chip-startup-nova-spectral-ai-raises-rs-21-crore-to-build-low-cost-image-sensors/articleshow/135216980.cms?from=mdr
[novaspectral.ai] Nova Spectral AI | Building the SWIR Image Sensor Chip | https://novaspectral.ai/
[viestories.com] Nova Spectral AI Raises $2.25 Million in Seed Round Led by Shastra VC | https://viestories.com/funding-alert/nova-spectral-ai-raises-225-million-in-seed-round-led-by-shastra-vc-12756859
[Dealroom.co] Nova Spectral AI raises $2.22M to build cheaper infrared sensors | https://app.dealroom.co/news/note/nova-spectral-ai-raises-2-22m-to-build-cheaper-infrared-sensors
[scoopearth.com] Nova Spectral AI Raises ₹21.15 Crore In Seed Funding Led By Shastra VC | https://scoopearth.com/nova-spectral-ai-secured-21-15-crore-in-a-seed-funding-round-led-by-shastra-vc/
[Lapaas Voice] Chip Startup Nova Spectral AI Raises ₹21 Crore to Develop Low-Cost Infrared Image Sensors | https://lapaasvoice.com/chip-startup-nova-spectral-ai-raises-%E2%82%B921-crore-to-develop-low-cost-infrared-image-sensors
[LinkedIn] Sanjeev Kumar, PhD (AI Agents) - Gen AI & NLP Leader | Incubate ideas into disruptive AI/ML Products | Monetize Agentic AI & LLM Research | Innovate in AI Agents, Search, Support & Enterprise AI | Lead Product, Eng., Solutions, R&D, Customer & Biz Dev | LinkedIn | https://www.linkedin.com/in/sanjeeku/
[CB Insights, retrieved 2026] Top SWIR Vision Systems Alternatives, Competitors | https://www.cbinsights.com/company/swir-vision-systems/alternatives-competitors
Articles about Nova Spectral AI
- Nova Spectral AI Raises $2.25 Million to Ship a 100x Cheaper Sensor — The deeptech startup is developing SWIR image sensors on standard wafers, targeting a leap from defense to consumer robotics.