cloud technology empowers us to seamlessly deploy our solutions across a multitude of cameras, adeptly managing vast volumes of audio and video feeds
New Delhi: Its AI-powered video analytics model JARVIS has helped law enforcement agencies crack many cases and now Staqu Technologies is providing retailers and manufacturers, ranging from Adani Power to WeWork, insights into real-time occupancy, efficient queue management and footfall analytics, says CEO Atul Rai.
Founded in 2015, Gurgaon-based artificial intelligence (AI) start-up Staqu provides image recognition, language-independent speaker identification, facial recognition and text processing, including sentiment analysis.
“Staqu’s vision is to transform every existing and video streaming device into intelligent analytical units. This transformation aims to first turn the existing dumb CCTV camera into a smart intelligence device empowered by Staqu’s proprietary AI platform JARVIS,” Rai told PTI.
Staqu is leveraging cloud through AWS. This has helped in easy deployment, on-demand processing and auto scale as per load, he said.
Excerpts from the interview:
Q: What are the key highlights in the journey of Staqu Technologies so far?
A: Since its establishment in 2015, Staqu Technologies has embarked on a pioneering path, addressing a wide range of client requirements around loss prevention and revenue enhancement. Our journey commenced with modest seed funding during which we developed our platform JARVIS, an AI-powered audio video analytics technology, for the Uttar Pradesh Police Department.
The platform extracted data from CCTV and other audio-video and language sources, resulting in a resounding success. Following this, we received invitations from several other state police departments, expanding our reach and impact.
Presently, we serve nine state departments, including Uttar Pradesh, Rajasthan, Bihar, Uttarakhand Punjab, and Haryana. Moreover, we extended our expertise to support the Bihar Election Commission in the critical task of vote counting where we used CCTV installed in the counting hall to perform optical character recognition (OCR) in wild to count votes directly from the EVM machine and report if there is any discrepancy with manual counting.
Buoyed by the success of our security solutions, we ventured into the in-revenue growth module for retail stores. This module mainly focuses on collecting insights from CCTV cameras which in turn is used in decision-making related to sales, operations and marketing.
These insights encompass distinct footfall analytics, real-time occupancy visualization, efficient queue management, and comprehensive customer journey analysis. We have provided these services to some of the leading retailers and manufacturing giants in the industry such as Adani Power, Croma, Starbucks, Caf Coffee Day and WeWork.
Q: What is the vision of the company and what are its mainstream offerings?
A: Staqu’s vision is to transform every existing and video streaming device into intelligent analytical units. This transformation aims to first turn the existing dumb CCTV camera into a smart intelligence device empowered by Staqu’s proprietary AI platform JARVIS. AI should make devices capable enough to automatically alert on security and safety breaches, and generate deep insights on operational processes and ultimately contributing to the advancement of a more sophisticated and forward-looking nation.
Our flagship offering, the JARVIS platform elevates traditional camera setups by infusing them with AI-powered audio and video analytics capabilities. This transformation not only enhances security and safety but also improves operational efficiency. It equips these cameras to deliver qualitative insights and valuable data to users, thereby redefining their utility and augmenting their overall contribution.
We have recently forayed into the Middle East region through a strategic partnership with a leading retail store chain, and have also established a new office in the United Kingdom.
Q: How do you see the transformative power of AI in various sectors and for various industries?
A: The immense potential of AI can be utilized for complex purposes across industries, based on individual requirements. AI has the capacity to revolutionize traditional processes and operations, leading to unprecedented advancements and efficiencies.
1. In retail, AI-powered analytics can provide personalized customer experiences, optimize inventory management, and predict consumer trends, leading to higher sales and customer satisfaction.
2. In manufacturing, AI-driven automation can streamline production processes, improve product quality, and enhance supply chain management. This results in increased productivity and competitiveness.
3. In security and surveillance, AI can identify and respond to threats in real-time, enhancing public safety and reducing risks.
4. In healthcare, AI can enhance diagnostics, drug discovery, and patient care, leading to improved medical outcomes and a higher quality of life. It can also aid in the analysis of vast medical datasets, facilitating research and innovation.
Q: For Staqu, how is AI enabling your products? How are you leveraging the power of AI to drive better outcomes for your customers? What challenges are you solving for your customers?
A: AI plays a pivotal role in enhancing Staqu’s products and delivering quantifiable results to our stakeholders. We leverage advanced image recognition and video analysis to provide customized AI solutions that are specially curated based on our client’s needs. Whether it’s providing actionable insights pertaining to customer behaviour in the retail industry, or predicting a potential equipment failure in manufacturing and optimizing supply chains in logistics, our tailor-made AI models and applications address diverse pain points to drive efficiency and revenue growth.
We excel in solving a range of challenges for our customers, including improving customer engagement through personalization, enhancing security and safety with real-time threat detection, optimizing resource allocation through demand forecasting, and automating operations to reduce costs.
Q: What are the new initiatives and innovations planned by the company?
A: We have recently introduced audio analytics capabilities within JARVIS, our flagship product. With this enhancement, we leverage the microphone audio from CCTV cameras to identify scenes and individuals. One notable application of this technology is in providing assistance during critical situations.
By incorporating these features, we can now detect SOS signals through both audio and video data. In instances where individuals require urgent assistance, they can simply vocalize the word ‘Help’ or make a waving gesture in front of the camera. This action triggers an immediate real-time alert.
Q: How have cloud technology and AWS enabled you in your operations?
A: The integration of cloud technology has significantly elevated our operational capabilities at Staqu.
In the past, this sector heavily relied on on-site hardware installations to conduct camera analytics, a process known for its time-consuming, expensive, and operationally demanding nature. Using cloud enabled us to do easy deployments, on-demand processing and auto scale as per load whereas for on-prem solutions addition of new features or new cameras would be restricted due to limited resources available.
Leveraging the power of the cloud through AWS, we unveiled JARVIS to end-users without necessitating any upfront capital investments. Our approach involves securely storing all feeds on the cloud in a manner that can be tailored to individual customers. This arrangement provides exclusive two-layer secure access for authorized stakeholders, ensuring the utmost confidentiality while eliminating the need for on-site hardware installations.
Furthermore, cloud technology empowers us to seamlessly deploy our solutions across a multitude of cameras, adeptly managing vast volumes of audio and video feeds. Simultaneously, it grants end-users the flexibility to incorporate additional features and analytics without being encumbered by hardware limitations.
With AWS, we were able to reduce the overall total cost of ownership (TCO) for our customers by using Amazon Simple Storage Service (Amazon S3), GPU servers including Amazon EC2 G4 for deploying machine learning and foundational models for image classification, object detection, Amazon EC2 G5 for cost-efficient training for moderately complex ML models and Amazon EC2 Inferentia (Inf1) for high-performance and low-cost machine learning inferencing.