India thesis

Complexity at scale is not a footnote. It is the design brief.

The world's most interesting technology problems are not always hiding in Silicon Valley. India concentrates linguistic diversity, institutional density, mobile-first behaviour, and millions of small businesses with thin digital tools.

Importing a product and translating the interface is not the same as asking what the product would be if it had been designed here from the start.

Voice, shared devices, interrupted connectivity, trust rituals, and bilingual documents are not edge cases. They are the centre.

If we solve for that centre with seriousness, we may also produce systems relevant to other complex markets.

From India

Scheduled languages and thousands of dialects in daily use
MSMEs running on WhatsApp, notebooks, and trusted relationships
Public complexity spanning schemes, documents, and eligibility rules
Trust requirements that cannot be bolted on after launch

Opportunity map

Domains where technology should work harder

Eight problem domains we treat as strategic - not because they are easy, but because India-scale complexity makes them worth solving exceptionally well.

Intelligence in every Indian language

Technology still treats many Indian languages as secondary. Dialect, code-mixing, and oral preference remain poorly served.

Why now: National language infrastructure is advancing, model capability is rising, and the cost of exclusion is clearer than ever.
India context: Scheduled languages, thousands of dialects, and mobile-first oral use create a design problem unique in scale.
Technology opportunity: Domain-aware speech and language systems, evaluation sets, and product patterns built for Indian interaction reality.

- What breaks first when an English-first assistant meets code-mixed speech?
- Which domains punish language failure most harshly?

Researching

Better technology for India’s small businesses

MSMEs lack tools that match their workflows, languages, margins, and trust requirements.

Why now: AI costs are falling while digital payments and messaging already sit at the centre of small-business life.
India context: Tens of millions of firms run on thin software, heavy human coordination, and WhatsApp-native operations.
Technology opportunity: Voice-first assistants, document understanding, and lightweight operational clarity.

- What is the smallest workflow worth automating well?
- How do we earn the right to hold business data?

Exploring

Trust in an AI-generated world

Synthetic media and generated documents raise the cost of believing digital claims.

Why now: Generation is cheap; verification UX and provenance practices are still immature.
India context: Population-scale platforms, linguistic diversity, and rapid mobile adoption amplify both harm and the need for practical trust tools.
Technology opportunity: Authenticity signals, verification workflows, and product patterns that surface uncertainty.

- What does consumer-legible provenance look like?
- Where must human review remain mandatory?

Researching

Education that adapts to every learner

Personal guidance and teacher support remain uneven across languages and contexts.

Why now: Adaptive systems are becoming cheaper, but evidence standards and vernacular quality still lag hype.
India context: Exam pressure, multilingual classrooms, device sharing, and teacher load shape what ‘personalisation’ can mean.
Technology opportunity: Teacher assistants, vernacular practice tools, and AI literacy with clear limits.

- How do we amplify teachers rather than sideline them?
- What evidence precedes outcome claims?

Exploring

Technology for public complexity

Citizens still struggle to discover, understand, and act on public information.

Why now: Digital public infrastructure expanded access - and also revealed new layers of navigational friction.
India context: Schemes, documents, and eligibility rules span languages, departments, and update cycles.
Technology opportunity: Cited information assistants and document explainers designed with strict accuracy discipline.

- How do we show sources and uncertainty by default?
- What should never be automated?

Exploring

Tools that make Indian professionals more capable

Document-heavy professional work wastes judgment time on retrieval and formatting.

Why now: Assistive models are strong enough to help - if grounded in local documents and audit needs.
India context: Bilingual files, portal workflows, and jurisdiction-specific practice break generic copilots.
Technology opportunity: Vertical productivity systems with retrieval, drafting support, and review trails.

- Which profession offers the cleanest first workflow?
- How do we prevent confident wrongness?

Exploring

Research infrastructure built from India

Discovery, evaluation, and collaboration tooling under-serve Indian research contexts.

Why now: Applied AI ambition is rising; infrastructure quality will determine seriousness.
India context: Local datasets, bilingual scholarship, and India-specific evaluation needs remain underbuilt.
Technology opportunity: Research assistants, dataset tooling, and transparent research communication systems.

- What open resources would compound fastest?
- How should public research label uncertainty?

Exploring

Interfaces that do not assume English literacy

Exclusion is designed into products that require literacy, bandwidth, and English comfort.

Why now: Voice models and multimodal interfaces make access-first design more feasible.
India context: Elderly users, low-literacy users, and disability communities are not edge cases at India scale.
Technology opportunity: Voice-first patterns, low-bandwidth modes, and accessibility as shared infrastructure.

- Who cannot complete the task today, and why?
- What does dignity look like in an interface?

Researching

Explore the Dream Index View problem domains

DREAMRESEARCHBUILDREALITY