Startup Deep Dive : QpiAI — the Rs 2.9 crore startup building India’s first full-stack quantum computer

In April 2025, on World Quantum Day, a Bengaluru company switched on QpiAI-Indus, a 25-qubit machine it calls India’s first full-stack quantum computer — and it did so on FY25 revenue of just ₹2.9 crore, a top line smaller than a single mid-sized restaurant’s. Three months later the same company, QpiAI India Private Limited, closed a $32 million Series A at a reported post-money valuation of $162 million, co-led by the Government of India’s National Quantum Mission.

That gap — between a nine-figure valuation and single-digit-crore revenue — is the whole QpiAI story. It is a deep-tech bet where the product is years from mass commercial demand, the government is both a customer and an investor, and the founder is asking backers to price a company on a roadmap rather than a receipt book. This is a look at what QpiAI actually sells, how it earns, who is funding it, and where the risk sits.

Quick facts

What QpiAI does

QpiAI builds quantum computers and the software that runs on them, and sells the combination to enterprises and governments as an optimisation and simulation tool. Its pitch is “full-stack”: it designs the qubit chip, the control and readout electronics, the cryogenics, the error-correction and compiler software, and the AI application layer, rather than assembling parts bought from others.

• Hardware: superconducting quantum computers. The flagship, QpiAI-Indus, uses 25 transmon (superconducting) qubits on a NISQ-class architecture (company and Wikipedia, April 2025).
• The stack around the chip: its own quantum processing unit, cryogenic system, control/readout electronics branded QpiAISense, and classical HPC integration (company documentation, 2025).
• Software and AI: platforms that pair quantum methods with generative AI for optimisation, simulation and machine learning — targeted at manufacturing, transportation, finance, pharma and materials discovery (company; TechCrunch, July 2025).
• Who buys: around 20 customers across India and the US, including the Indian government, as company-stated to TechCrunch (July 2025).

The origin

QpiAI began as a bootstrapped project in 2019, started by Dr Nagendra Nagaraja after a semiconductor and wireless career spanning NVIDIA, Qualcomm and Lucent Microelectronics. He holds an engineering degree from Bangalore University, a master’s from the Illinois Institute of Technology and a PhD tied to wireless, machine learning and AI; his profile lists 27 of 30 patents granted across wireless, security, multimedia and AI.

The founding insight was not quantum for its own sake. Nagaraja had already co-founded AlphaICs in 2016 with veteran chip architect Vinod Dham to build AI accelerators, so he arrived at QpiAI from the AI-hardware side. His argument, made repeatedly since, is that classical AI hits a wall on hard optimisation and simulation problems, and that quantum processors are the way to make those workloads tractable — “quantum can really make sure AI is robust,” as he put it to TechCrunch. That framing, quantum in service of AI rather than as a science project, is why the company is called QpiAI and not simply a quantum-hardware brand.

The struggle years

The hard part of QpiAI’s history is structural, not dramatic: for its first five years it was building physics-grade hardware in a country with almost no domestic supply chain for it, and with very little revenue to show. The company was bootstrapped from 2019, and did not raise meaningful external capital until 2024.

• No local supply chain: the founder has said roughly 20% of the foundry’s inputs are still imported even now (qnews.io, August 2026); in the early years, sourcing superconducting-grade materials and cryogenics domestically was harder still.
• Five years bootstrapped: the first outside round, a Pre-Series A of $6.5 million, only closed in June 2024 — nearly five years after incorporation (TechCrunch, July 2025).
• Revenue far below the science: even in FY25, after the flagship launch, revenue was ₹2.9 crore (Tracxn / Tofler) — a rounding error against the cost of building quantum hardware, which the company has had to fund from equity and grants rather than sales.
• Talent scarcity: the company needs PhD-level quantum and cryogenics engineers, a thin pool in India; its stated plan to reach 60-plus PhDs on staff is itself a hiring gamble (TechCrunch, July 2025).

The turning point

The turning point is a pair of events three months apart in 2025. On 14 April 2025, World Quantum Day, QpiAI launched QpiAI-Indus, a 25-qubit superconducting machine it billed as India’s first full-stack quantum computer — a hardware milestone that gave the company something concrete to sell and to be valued on. Then, on 15–16 July 2025, it closed the round that changed its profile.

• Before: a bootstrapped-then-lightly-funded startup on a ~$30 million valuation at its June 2024 Pre-Series A (TechCrunch).
• After: a $32 million all-equity Series A at a reported $162 million post-money valuation — roughly a five-fold step up in about a year — co-led by the National Quantum Mission and Avataar Ventures (TechCrunch; Business Standard, July 2025).
• Why it matters: the government co-leading an equity round, not just handing out a grant, is the signal. It made QpiAI the National Quantum Mission’s most prominent hardware bet and, by valuation, India’s best-funded quantum company (TechCrunch, July 2025).

The money behind it

QpiAI’s capital has come in two clear equity rounds plus government grant money, and the headline totals differ by source depending on what is counted.

• Pre-Series A — $6.5 million (June 2024): led by YourNest and SIDBI Venture Capital (SVCL), with angels including ex-Cognizant CEO Lakshmi Narayanan and the founders of Quick Heal, at a roughly $30 million valuation (TechCrunch).
• Series A — $32 million (July 2025): all-equity, co-led by the National Quantum Mission (via the Department of Science and Technology) and Avataar Ventures, at a reported $162 million post-money valuation (TechCrunch; Business Standard).
• National Quantum Mission context: QpiAI is one of eight startups selected under India’s roughly $750 million National Quantum Mission, each eligible for an initial grant of up to $3.5 million (TechCrunch / reporting, July 2025).
• Total raised — figures vary: Inc42 puts cumulative funding at about $38.5 million, TechCrunch at “nearly $39 million” of equity, and Tracxn at up to $54.1 million across nine rounds when grants and tranches are included (Inc42; TechCrunch; Tracxn, 2026). Treat the equity total as high-$30-millions and the wider figure as source-dependent.

How it makes money

QpiAI’s revenue model is broader than selling finished quantum computers, which very few customers can yet absorb. The money comes, or is meant to come, from several layers of the same stack.

• Systems and access: selling or providing access to quantum machines (Indus and successors) to enterprises, research institutions and government labs.
• Software and optimisation services: quantum-plus-AI platforms sold into manufacturing, finance, pharma, materials and logistics for optimisation, simulation and drug/material discovery (company; TechCrunch).
• Chip fabrication: its Bengaluru foundry, whose second phase was inaugurated on 17 August 2026, runs an eight-inch line capable of chips up to 128 qubits and does lithography, etching, assembly and packaging in-house — positioning QpiAI to make chips, not just buy them (qnews.io, August 2026).
• Where the margin is meant to sit: the founder claims roughly 60% gross margins and 20–30% net margins with about 80% of hardware assembled in-house (company-stated, TechCrunch). Those are founder figures, not audited, and sit uneasily against the loss-making picture in the filings below.
• The part people get wrong: this is not yet a product business at scale. On FY25 revenue of ₹2.9 crore, the company is effectively pre-commercial, funding itself on equity and grants while it builds toward a stated ₹100 crore (about $10.4 million) revenue goal.

The numbers

The reported financials are small and, importantly, they conflict with the founder’s margin claims — a tension worth stating plainly rather than smoothing over.

• Revenue: ₹2.0 crore in FY24 rising to ₹2.9 crore in FY25, a 43% increase, per Tracxn and Tofler drawing on MCA filings (year ending 31 March 2025).
• Profitability: filings-based sources indicate substantial operating losses in FY25, with EBITDA down heavily year on year even as book net worth rose (Tracxn / Tofler) — consistent with a company spending heavily on hardware R&D.
• Founder counter-claim: Nagaraja has told TechCrunch the company has run positive EBITDA “for three years” with ~60% gross and 20–30% net margins. This is company-stated and does not reconcile with the filings; on the evidence, treat the audited-filings picture as the more reliable one and the founder’s margins as unverified.
• Target: ₹100 crore (about $10.4 million at $1 ≈ ₹96.0) in annual revenue, a goal the founder has attached to 2026 — roughly a 35x jump from FY25 (company-stated, TechCrunch).

Where the money comes from

QpiAI’s demand is split across geographies and sectors, and the surprise is how central the government is on both sides of the ledger — as buyer and as backer.

• Geography: around 20 customers across India and the US, with subsidiaries in the US and Finland to sit close to Western research and talent (TechCrunch, July 2025).
• Sectors: optimisation and simulation demand aimed at manufacturing, transportation, finance, pharma and materials/drug discovery (company; TechCrunch).
• The government is the anchor: the Indian government is named as a customer and is also a co-lead investor via the National Quantum Mission — an unusually tight buyer-plus-backer relationship (TechCrunch).
• Expansion markets: the company has flagged Singapore, the Middle East and Southeast Asia, and lists the US as an export market for foundry output (TechCrunch; qnews.io).
• The hardware ladder that is meant to grow revenue: a product roadmap running QVidya (8 qubits), Yukti (9), Indus (25), Kaveri (64), Ganges (128) and a long-range Everest (up to ~1,000 qubits), with a stated ambition of 10,000 physical qubits of fabrication capacity by 2027 (Wikipedia; qnews.io, August 2026).

The risks

• Valuation vs revenue mismatch: a reported $162 million valuation on ₹2.9 crore of FY25 revenue prices the company almost entirely on a roadmap. If commercial quantum demand arrives slower than the roadmap assumes — and NISQ-era machines have limited proven enterprise use cases — the down-round risk at the next raise is real.
• Disclosure gap on profitability: the founder’s claim of positive EBITDA and healthy margins does not match filings-based sources showing operating losses. Whichever is right, the mismatch itself is a governance and diligence risk for an investor pricing the company on its economics.
• Supply chain and capital intensity: about 20% of foundry inputs are still imported, and building and expanding an eight-inch quantum-chip line is capital-hungry — an initial $20–25 million outlay plus a planned $10–15 million more (qnews.io, August 2026). A funding pause could stall the hardware roadmap that the valuation rests on.
• Talent and execution: scaling to 60-plus PhDs and hundreds of staff in a shallow domestic quantum talent pool, while hitting a 35x revenue jump to ₹100 crore, is a demanding double execution bet.
• Concentration on the government: the National Quantum Mission being both a lead investor and a customer helps today, but it concentrates QpiAI’s fate in one policy programme; a shift in mission priorities or budgets would hit both funding and demand at once.

The takeaway

The transferable lesson from QpiAI is about sequencing in deep tech. Nagaraja spent five years bootstrapping and building an integrated stack before raising serious money, and then used a single credible hardware launch — QpiAI-Indus on World Quantum Day — to reprice the company roughly five-fold in a year. In frontier hardware, the milestone that unlocks capital is a working machine, not a pitch deck; the capital follows proof, and the proof has to be built first, often on almost no revenue. The open question QpiAI now has to answer is the one every deep-tech story eventually faces: can it convert a physics achievement into a business before the money priced on its promise runs out.

Frequently asked questions

What is QpiAI and what does it build?

QpiAI (legal entity QpiAI India Private Limited) is a Bengaluru deep-tech company founded in 2019 that builds full-stack superconducting quantum computers and pairs them with AI software for optimisation and simulation. Its flagship machine, QpiAI-Indus, is a 25-qubit system launched on 14 April 2025.

Who founded QpiAI?

Dr Nagendra Nagaraja, its Founder, CEO and Chairman. He previously worked at NVIDIA, Qualcomm and Lucent Microelectronics and co-founded the AI-chip startup AlphaICs in 2016 before starting QpiAI.

How much has QpiAI raised and at what valuation?

It raised a $32 million all-equity Series A in July 2025, co-led by India’s National Quantum Mission and Avataar Ventures, at a reported $162 million post-money valuation, after a $6.5 million Pre-Series A in June 2024. Cumulative funding is reported at roughly $38.5–39 million of equity, or up to $54.1 million across all rounds including grants, depending on the source.

Is QpiAI profitable?

The company reported ₹2.9 crore of revenue in FY25 (up 43% from ₹2.0 crore in FY24) and, on filings-based sources, was loss-making at the operating level. The founder has separately claimed positive EBITDA and 20–30% net margins, but that is company-stated and does not match the filings.

What is the National Quantum Mission’s role in QpiAI?

The National Quantum Mission, part of India’s roughly $750 million quantum programme under the Department of Science and Technology, is both a co-lead investor in QpiAI’s Series A and a customer. QpiAI is one of eight startups the mission has selected, each eligible for an initial grant of up to $3.5 million.

Sources

Figures are as of September 2026. Currency converted at $1 ≈ ₹96.0 as of 18 September 2026 (Trading Economics).

• TechCrunch, “India eyes global quantum computer push — and QpiAI is its chosen vehicle” (July 2025)
• Business Standard, “QpiAI raises $32 million from Avataar, NQM to build quantum computers” (July 2025)
• BW Businessworld, “Inside QpiAI: Can the NQM-backed startup become India’s first quantum unicorn?” (2025)
• qnews.io, “QpiAI opens quantum chip foundry in Bengaluru” (August 2026)
• Wikipedia, “QpiAI-Indus” and “Nagendra Nagaraja” (2025–2026)
• Tracxn, QpiAI / QpiAI India Private Limited company profile (2026)
• Tofler and Inc42, QpiAI India Private Limited financials, from MCA filings (2026)
• Zauba Corp / Instafinancials, QpiAI India Private Limited registration (CIN U72900KA2019PTC129453)

Found an error? Write to us and we’ll correct it in the open, dated, on the piece.