Analytics India Magazine caught up with Umesh, Co-Founder & Chief Operations Officer at CamCom to understand the underlying technology that focuses on AI, and how it is providing hyper-intelligent quality assurance.
Founded by Ajith Nayar, Mahesh Subramanian and Umesh in 2017, CamCom is an AI platform for automated quality assurance. It provides contactless defect and damage assessment solutions in the automotive industry. The new solution is claimed to provide contactless, scalable, accurate and cost-effective assessment. The platform is poised to be the imperative solution in the post COVID world to future proof enterprise quality processes.
AI To Make The Product Game Strong
According to Umesh, automotive defect and damage assessment are the flagship products. The hyper-intelligent quality assurance platform at CamCom aims to help to drive efficiencies across the automotive value chain, from various sectors such as manufacturing, motor insurance, used cars sector, among others.
The AI solution is claimed to eliminate human subjectivity from the quality control process entirely with the help of using a computer vision stack. Bespoke rigs with high-end machine-vision cameras for production or logistics are deployed on the mobile app for after-market solutions.
In the motor insurance sector, CamCom provides self-inspection models for spot settlement of claims and instant premium calculation for break-in policy renewals. These are then utilised through integration with insurance company apps for customers, surveyor apps or web app links sent by SMS generating an accurate report for consumers.
In the used cars sector, CamCom provides an automatic visual health check report for a true value assessment yielding a single source of truth visual audit trail for service centres to enable upsell opportunities. Umesh further mentioned another product that is an add-on to the flagship offering, known as Automatic Quality Control.
How Is It Differentiated From Other Products In The Market
To this, Umesh replied that the core differentiator lies in the fact that CamCom goes beyond other QC companies that leave the heavy lifting to the enterprises. The key focus is to understand the business of the enterprise and recognise the gaps that need to be filled.
He added, “We recalibrate and integrate our platform into the needs of the enterprise while ensuring seamlessness in the workflow and without compromising on the quality assessment and checks.”
Use of AI & ML @ CamCom
Top of the list of machine learning that is used at CamCom is advanced computational methods that ultimately show patterns and inferences. Umesh said that as CamCom deals with micro damages on images, it is essential for the machine to identify and differentiate a pattern (scratch or dent) from a reflection that may appear.
Some of the important techniques in AI and computer vision that the company uses are object detection, object tracking and image reconstruction, image classification, instance and semantic segmentation.
Data mining is another aspect that is derived from the platform. Umesh shared that with data mining, it is possible to pinpoint where the product lacks or fails and which vendor is responsible for it.
Core Tech Stack
The core technology stack at CamCom lies in REST APIs, Python, NGINX Gunicorn, Public CDNs, MySQL, Tensorflow, PyTorch, Ubuntu Farm, Nvidia GPUs.
- At the foundation level, CamCom uses Intel Architecture with Nvidia GPUs running Ubuntu Operating System.
- The AI Layer comprises Tensorflow and PyTorch.
- CamCom has built a micro-services architecture that comprises API Stack, queuing services, assessment services, reporting services, web app services, and container services.
- On the presentation layer, the company provides APIs, dashboards and alerts to the customers.
Initially, CamCom raised an angel funding of about $150,000, and revenues are expected to hit between the range of 750,000 to one million US dollars by the end of this year. In the coming years, the company is aiming to go global.
“We aim to provide this solution on the international stage as the ‘go-to’ company for defect and damage solutions. Although this goal is ambitious, we plan to grow steadily but surely using a partner-driven approach. Post-COVID and in the world of the new normal, partner networks will prove to be a key ingredient in our success,” said Umesh on a concluding note.
If you loved this story, do join our Telegram Community.
Also, you can write for us and be one of the 500+ experts who have contributed stories at AIM. Share your nominations here.
A Technical Journalist who loves writing about Machine Learning and Artificial Intelligence. A lover of music, writing and learning something out of the box. Contact: firstname.lastname@example.org