About Smriti AI Labs
We engineer intelligence for the real world.
Why we exist
What has to happen between a model producing an output and a person being able to use it?
Smriti AI Labs was founded around a shared ambition: to build AI systems that move beyond demonstrations and become useful parts of the real world. Our work starts with that practical question.
For a business, that means connecting intelligence with its knowledge, software and workflows. For an automotive application, it means designing and testing against the environment the system must face. For an AI product, it means engineering the infrastructure that allows the model to run as the application needs.
Smriti brings these challenges into one company.
Our lead business is the development of enterprise AI products and agentic platforms for India. We build on Google Cloud and AWS, combining established AI technologies with our own voice models, AI harnesses and product engineering.
Alongside these platforms, our automotive practice designs, tests and refines NVIDIA Hyperion-based ADAS solutions. Our accelerated-inference practice engineers NVIDIA GPU-based model execution. Both are part of our work today.
Our longer-term ambitions include research toward Level 5 autonomy and the development of dedicated low-latency and GPU cloud infrastructure.
India is not an afterthought in our roadmap. It is the starting point for the systems we build.
We bring models, software and infrastructure together—and take responsibility for the engineering between them.
Principles
What guides the work.
Systems, not demos
A system succeeds when it does its job in the place it has to run — not when it looks good in a presentation.
Ownership made explicit
We say which layers a platform provides and which Smriti engineers. We never present Google, AWS or NVIDIA technology as our own.
Evidence before claims
Project records carry an honest stage. Results are published only when evidence supports them, and illustrative material is labelled as such.
People stay in control
Where a decision needs a person, the system is designed to involve one — and to record what was decided.
Built for India
India is not an afterthought.
We build for businesses in India from the first design decision, not as a localisation step at the end. Products are shaped around the organisations that will use them and the workflows they actually run. Our automotive work is designed and evaluated for Indian automotive applications and the operating conditions they must face.
Global platforms give us capable foundations. Engineering them for the Indian context is our work — including the languages an application must handle, which we define and evaluate for each application rather than claim in general.
Our work today
Three current businesses.
Scalable AI products and agentic platforms for businesses in India, built on Google Cloud and AWS, alongside our own voice models and AI harnesses.
What we do today
- Agentic platforms and enterprise applications
- Gemini-based systems on Google Cloud
- Amazon Bedrock-based platforms on AWS
- Smriti voice models and voice-enabled systems
- AI harnesses, evaluation and deployment
Technologies we build with
- Google Cloud
- Gemini Enterprise
- Gemini models
- Agent Development Kit (ADK)
- Cloud Text-to-Speech
- AWS
- Amazon Bedrock
Design, testing and refinement of ADAS solutions based on NVIDIA Hyperion and relevant NVIDIA DRIVE technologies, for Indian automotive applications.
What we do today
- ADAS solution design for intended operating conditions
- AI software and model adaptation
- Testing, evaluation and evidence-led refinement
Technologies we build with
- NVIDIA Hyperion
- NVIDIA DRIVE
Engineering of NVIDIA GPU-based inference systems: model deployment, serving and performance optimisation for AI applications.
What we do today
- Model deployment and serving on NVIDIA GPUs
- Workload analysis and performance optimisation
- Evaluation of response time, throughput, resource use and quality
Technologies we build with
- NVIDIA GPUs
- NVIDIA inference software, selected per workload
Longer-term ambitions
Where we intend to go.
Kept apart from our current work everywhere on this site, and labelled as ambitions.
DirectionAutomotive Intelligence
Research toward Level 5 autonomy
A long-term research ambition within our automotive practice. It is not a capability we claim today, and our current ADAS work is not Level 5.
DirectionInfrastructure direction
Dedicated low-latency cloud infrastructure
Infrastructure designed around the latency needs of the AI systems we build. A future direction, not a current service.
DirectionInfrastructure direction
Smriti-owned GPU cloud capacity
Future GPU capacity of our own. We do not operate a self-service GPU cloud today; our current inference work is engineering on NVIDIA GPU-based systems.
Founders
Led by its co-founders.
Krishna
Co-Founder
Shreya
Co-Founder
Start a conversation