Cactus Technology Solutions has been selected to provide artificial intelligence and machine learning expertise to government digital initiatives, focusing on secure and scalable AI solutions.
Key Points
- Cactus Technology Solutions will provide AI and machine learning expertise to government digital initiatives.
- The company will focus on building secure and scalable AI systems tailored to real-world needs.
- Cactus will work with NeGD to enable end-to-end AI implementations, including data science and model development.
- The company plans to implement guardrails, including a moderation engine, to ensure responsible AI usage.
The National e-Governance Division (NeGD), under the IT ministry, has empanelled Cactus Technology Solutions Private Limited (CACTUS) to provide specialised artificial intelligence/machine learning expertise to design and deploy scalable AI solutions for government-led digital initiatives.
Secure And Scalable AI Systems
The IT company will focus on building systems that are secure, scalable, and grounded in real-world needs, as ministries and departments continue to adopt AI.
"As part of this engagement, CACTUS will work closely with NeGD to enable end-to-end AI implementations, spanning data science, model development, and deployment through the provision of specialised talent," according to a statement released by CACTUS on Tuesday.
AI Applications In Law Enforcement
In an interview with PTI, Akhilesh Ayer, CEO of the company, said that CACTUS could build systems as per the requirements, creating data, processing, and app layers.
"For example, with the new legal laws (Bharatiya Nyaya Sanhita), we can create an app for law enforcement through which they can recall a specific clause at the click of a button. This could be available in all languages across all states," said Ayer.
Ensuring Responsible AI Usage
To ensure that AI systems do not falter, CACTUS plans to implement several guardrails, including deploying a unified moderation engine that can scan both inputs and outputs across text, images, and metadata.
The engine can detect, for example, hate/abusive language, sexual/explicit content, and violence/self-harm signals.
"Every aspect of the systems will have a layer of responsible AI," said Ayer.
