Indian MSMEs: AI remains stuck at the PoC door, as in Singapore

In plain terms. India has over 60 million MSMEs (Micro/Small/Medium Enterprises) that drive the country's economy. YourStory (July 12, 2026) documents that the vast majority do not use AI—not by ideology, but by barrier to entry. The pattern is the same as in Singapore: the PoC stalls before production.

Context

Two YourStory articles from July 12 frame the scene: Fynd (Ragini Varma) on AI adoption for autonomous commerce, and Mintoak (Raman Khanduja) reminding that the Indian QR revolution (UPI, PhonePe, GPay) is just the beginning for MSMEs. The convergent message: digital foundations exist (payments, KYC, GST), but AI remains out of reach—not the model, the infrastructure to apply it.

The Data

• 63.4 M MSMEs registered in India (Ministry of MSME, data as of their date).
• Contribution: ~30% of GDP, ~45% of exports.
• Operational AI adoption: figure absent from the articles, but presented as marginal—the cases cited (Fynd) remain exceptions.

Analysis

What causes the PoC-to-production failure for Indian MSMEs mirrors the Singaporean pattern identified in our previous publications:

• Data quality—MSMEs often operate on Excel + WhatsApp. AI does not compensate for the lack of schema.
• Unit cost—GPT-class inference is still too expensive compared to an average MSME ticket. Indian models (Bhashini, Sarvam) are beginning to close this gap.
• Skills—the pool of integrators who know how to build a profitable AI workflow for an MSME remains rare.

The difference with Singapore is the market scale—60 million potential customers justify building the infrastructure. Fynd and Mintoak are betting on this volume.

Scenarios

Base (55%): 5-10% of Indian MSMEs will use an AI tool by mid-2027, but for peripheral tasks (content, customer responses). Optimistic (25%): Low-cost Indian models (< $0.001/req) unlock the segment for finance + logistics workflows. Pessimistic (20%): PoC remains PoC, the gap between AI-native unicorns and MSMEs widens.

Implications for the Carrier

The opportunity is not in the model; it is in vertical integration: who builds the equivalent of Toast (US restaurant) but for an Indian MSME sector with AI embedded from the start?

Risks

Ongoing regulation (IndiaAI, DPDP Act), rural connectivity infrastructure, dependence on foreign models.

So What

The story of Indian digital is one of public infrastructure (UPI, Aadhaar, ONDC) that unlocked the private sector. AI will follow the same path—or not at all. Watch for the announcement of an Indian public AI layer equivalent to UPI, or conversely, the complete privatization of the layer.