The International Labour Organisation (ILO) estimates that workplace safety risks result in economic losses of up to 5.4 per cent of global GDP annually. In response, businesses have implemented over one billion CCTV cameras and employed hundreds of thousands of security personnel and risk managers to monitor footage and mitigate risks.
However, this process is slow, labour-intensive, costly, and prone to inaccuracies.
“Video is the richest data source, but existing solutions are too rigid, fragmented, impossible to customise, and challenging to scale. This makes it difficult for business owners to maximise the video data and manpower they have on hand to make effective decisions. We see a much better way,” says Natalie Doran, CEO and Co-Founder of Lytehouse, in an email interview with e27.
Lytehouse offers real-time video intelligence that automates risk management by transforming any CCTV camera feed into virtual co-workers, enhancing efficiency and reducing costs. Its AI-powered solution applies advanced computer vision (CV) models, machine learning (ML) algorithms, and generative AI (GenAI) agents to existing cameras, enabling businesses to capture key data points, gain real-time insights, and automate manual tasks for improved decision-making.
Using Large Language Models (LLM), Lytehouse makes video insights easily queryable in natural language. Business owners can request specific actions, such as notifying managers if workers use mobile phones near machinery, to detect risk events and generate reports.
The scalable system tailors GenAI agents to specific industries and tasks, allowing companies to address safety and security risks without replacing their existing camera infrastructure.
Also Read: Gen AI in banking: How to ensure a successful transformation for an age-old industry
“Our approach of effectively giving business owners a fleet of digital workers that can be deployed directly onto their cameras is the biggest difference compared to the alternatives, as it provides enormous flexibility in what these agents should be looking for within various video streams without locking our customers into configuration hell,” Doran says.
The launch of Lytehouse’s security agent marked a significant milestone for the company. Additionally, Lytehouse secured a major partnership with one of Africa’s largest CCTV distributors, providing access to hundreds of thousands of cameras across the region.
Lytehouse’s journey began as a personal mission for security, driven by real-life experiences. Jean-Vicente De Carvalho, CTO and Co-Founder of Lytehouse, comes from a family of small business owners in South Africa who endured 37 instances of armed robberies, kidnappings, and bombings.
In 2018, after his father was held at gunpoint despite the presence of extensive security measures, De Carvalho realised the urgent need for a more effective solution.
“I met Jean in the Entrepreneur First programme in Singapore, and we did a deep dive into security and risk management. We realised there was a huge opportunity to give existing manpower the superpowers to respond effectively to risk, and beyond this, we could empower businesses with the right data to be more proactive and preventative. And so, Lytehouse was born,” Doran explains.
Also Read: Singapore surpasses US in AI investment: Study
“We have evolved a lot since then, and so have the technology capabilities to address these global pain points. We’re proud to say we have assembled a rockstar team driven by the passion to make a tangible difference and have the backing of incredible investors, government departments, and tech giants.”
Lytehouse is part of the inaugural cohort of AI startups graduating from the Google for Startups Accelerator: AI First Singapore, which ran from April to August 2024.
According to Doran, collaborating with Google Cloud has been a transformative experience for the company. The announcement of Gemini 1.5 Pro and Flash models at Google I/O, with their extended context windows and multimodal support, marked a pivotal moment for the company.
By working closely with Google Cloud on the Vertex AI platform, Lytehouse accelerated the development of new product lines. Beyond technical support, Google Cloud offered significant business opportunities, such as featuring Lytehouse on the Google Cloud Marketplace and connecting them with customers across Southeast Asia.
“We’re crafting a future where technology expands human potential, replacing tasks, not jobs, for more meaningful work experiences. To achieve this, we are building the world’s most performant, cost-effective, secure, and extensible intelligence engine to annotate the world’s video data and make it useful,” Doran closes.
“We believe that by the end of 2025, thanks to our powerful intelligence engine, there will be a use case for Lytehouse for every camera in the world – and by leveraging our strategic partners, we will have the reach to attain a large proportion of them.”
—
Image Credit: Lytehouse
The post Lytehouse improves workplace safety with AI-based video intelligence, teams up with Google Cloud appeared first on e27.