Y Combinator’s Summer 2024 Request for Startups identified spatial computing as a promising area for ambitious companies. The application window has closed, but the underlying opportunity remains relevant: software that understands places, objects, movement and human context can solve problems that conventional 2D apps cannot.
For founders in India, the strongest opportunities are unlikely to be metaverse replicas. They are more likely to be practical systems for industrial training, field service, healthcare, construction, logistics, retail and education—markets where spatial information directly affects cost, safety or productivity.
What spatial computing means for builders
Spatial computing is a broad category covering software and hardware that sense, map or augment the physical world. It may include:
- Augmented and mixed reality: Digital instructions, data or objects overlaid on a user’s surroundings.
- Virtual reality: Fully simulated environments for training, design, therapy or entertainment.
- Computer vision and 3D perception: Systems that identify objects, surfaces, depth and human activity.
- Geospatial computing: Location-aware applications using maps, satellite data, sensors and positioning.
- Spatial interfaces: Interaction through gaze, hand gestures, voice, movement or room-scale context.
- Digital twins: Software representations of factories, buildings, infrastructure or equipment that can be queried and updated.
The category becomes commercially meaningful when spatial input leads to a measurable decision or action. A technician finishing a repair faster, a surgeon rehearsing a procedure, or a warehouse reducing picking errors is a clearer value proposition than “an immersive experience”.
Why the opportunity is different in India
India offers unusually varied environments for spatial products: dense cities, multilingual users, large industrial networks, distributed healthcare and extensive infrastructure projects. This creates difficult operating conditions—but also strong product moats for teams that solve them.
A startup might build a visual inspection tool that works in low light, a training product that runs on affordable Android devices, or a construction workflow that remains useful with intermittent connectivity. Local language support, device affordability and deployment through enterprise partners can matter as much as the underlying 3D technology.
Founders should also account for privacy, consent and data governance. Spatial products can capture faces, voices, homes, workplaces and sensitive industrial layouts. Minimise collection, secure stored scans, define retention policies and make it clear when recording or analysis is active.
What YC’s thesis means for a startup pitch
YC does not require a company to use a particular headset or computer-vision model. Its recurring interest is in teams that identify a large problem, build quickly and demonstrate evidence that users want the solution. A spatial-computing pitch should therefore answer five questions:
- Who has the problem? Name the operator, buyer and economic owner—not just “users”.
- What spatial signal matters? Explain which location, depth, object or movement data changes the workflow.
- Why is a spatial interface necessary? Show why a phone, desktop dashboard or ordinary video call is insufficient.
- What improves? Quantify time saved, errors avoided, revenue created or risk reduced.
- Why can your team win? Highlight domain access, technical expertise, proprietary data or distribution.
A short demo is useful, but it should prove the workflow rather than merely display visual effects. Capture a before-and-after task, show failure cases and state what the prototype cannot yet do.
Product opportunities worth testing
Industrial operations and field service
Technicians can receive step-by-step visual guidance, identify components through a camera and escalate difficult repairs with remote experts. The buyer may be an industrial company, equipment manufacturer or service contractor. Measure first-time fix rate, training time and mean time to repair.
Construction and infrastructure
Teams can compare building plans with on-site conditions, detect deviations and coordinate contractors around a shared model. Integrations with existing BIM, project-management and survey tools are essential; a standalone viewer is rarely enough.
Healthcare and medical education
Spatial tools can support anatomy learning, surgical planning, rehabilitation and clinical training. Start with low-risk workflows and validate with qualified professionals. Clinical claims, patient data and device certification can materially extend sales cycles.
Logistics and retail
Computer vision can support warehouse picking, shelf compliance, store navigation and product visualisation. In India, test the system across varied store layouts and lighting conditions rather than relying on a controlled demonstration environment.
Skills and vocational training
Immersive simulations can help workers practise hazardous or expensive procedures. The product must connect training performance to an employer’s hiring, certification or safety process; otherwise it may remain an attractive but discretionary purchase.
A practical validation plan
Start with one workflow and one buyer. Interview 15–20 target users, observe the task in its real setting and document the current workaround. Collect baseline data before building a polished experience.
Then create the smallest useful prototype. A phone camera, web viewer or lightweight headset may be enough. Teams can use rapid AI prototyping services for startups to test perception, retrieval or assistant features before investing in custom hardware.
Evaluate the prototype against operational metrics:
- Task completion time
- Error or rework rate
- Training hours required
- Accuracy across devices and environments
- Network and battery consumption
- Willingness to pay or signed pilot commitments
For voice-driven field workflows, multilingual interaction may be important. A prototype can draw on lessons from building multilingual chatbots for Indian startups, while keeping spatial recognition and language interaction as separate components that can be tested independently.
Technical and commercial architecture
A practical stack often combines mobile or headset clients, computer vision, 3D or geospatial data, a backend for synchronisation and analytics, and an admin interface for supervisors. Choose the device only after understanding the environment. High-end headsets may provide better tracking but can restrict adoption; phones and tablets improve reach but may reduce immersion.
Keep latency-sensitive perception on-device where possible. Send only the data needed for synchronisation, auditing or model improvement. Test under Indian network conditions, including unreliable connectivity and shared devices. For model-heavy features, benchmark infrastructure cost per active user rather than relying on a laboratory demo. A broader tech stack guide for AI startups can help structure those decisions.
Spatial products also need strong workflow automation. If an inspection produces a report, a repair ticket or an escalation, integrate that action instead of stopping at visual recognition. Patterns from AI workflow automation for high-growth startups are relevant here: define triggers, human approvals, audit trails and failure recovery.
How to prepare a stronger application or investor conversation
Although the Summer 2024 YC process is closed, the same evidence strengthens applications, pilots and fundraising in 2026. Prepare:
- A one-sentence description of the customer, workflow and outcome.
- A 60–90 second product demonstration in a real environment.
- Baseline and pilot metrics, including unsuccessful tests.
- A clear explanation of hardware, software and deployment requirements.
- Customer references, paid pilots or credible letters of intent.
- A plan for distribution and expansion beyond one site or enterprise.
Avoid inflated market claims based only on the size of AR, VR or AI. Show the initial wedge, the budget that funds it and the adjacent workflows that could support expansion.
Bottom line
The durable lesson from YC’s spatial-computing request is not that every startup needs a headset. It is that physical context can unlock valuable software when it improves a real job. Indian founders should prioritise constrained workflows, measurable outcomes, affordable deployment and responsible handling of spatial data. Build the narrowest useful product, test it in messy real-world conditions and let customer evidence—not novelty—determine whether the company deserves to scale.