Computer vision is a type of AI that involves developing algorithms, models and systems that can process images and videos, recognize objects and patterns, and extract useful information from visual data. We are then able to use these insights to trigger actions or automated responses resulting in tangible business benefits.

Inveniam provides an end-to-end managed solution across the computer vision lifecycle from use case definition to building, deploying and managing complex computer vision and machine learning models. We provide a web-based software as a service platform where customers can easily upload and review input data, track status across the data pipeline, as well as review output data, dashboards and key results.

Whether you require hundreds or millions of images to be processed every month we have you covered. Our platform is designed to scale, and regardless of your needs we're equipped to support you. For specific metrics or to discuss the extent of our processing capabilities, please contact our Sales team.

Our platform supports a wide range of visual data, including images, videos and sensor data, as well as meta data. From high-resolution images to streaming videos, we provide comprehensive analysis and predictions for diverse use cases.

We offer both solutions. We have pre-packaged, off-the-shelf use cases that cater to common industry needs such as shared infrastructure access. We also provide custom solutions, tailoring or developed completely new models to fit your specific requirements and to deliver the best results.

We follow the CrispDM methodology, a proven industry-standard for data mining. Our process starts by thoroughly understanding your business needs. We then gather relevant data, analyse it to derive patterns, and experiment and train predictive models leveraging the latest AI architectures suited to each case. Every step is tailored to ensure that the AI model aligns with your specific objectives.

Choosing our models offers several advantages:  Faster outcomes: Our pre-trained and customised models expedite the implementation process.  Lower Total Cost of Ownership (TCO): Leveraging our team eliminates many upfront costs and reduces maintenance overhead, without the need to invest in or scale up your own specialist AI team. Expertise: Our team of experts are well-versed in the latest AI trends and technologies.  Innovation: We continuously update our solutions to stay at the forefront of technological advancements. 

Yes, our computer vision and machine learning models support real-time analysis of visual data, providing your team with immediate insights and predictions. Real time data analysis is only available to customers integrated with Inveniams’s API.

We employ a continuous improvement approach, monitoring live data and assessing the accuracy of our predictions. Our active learning strategies also mean that our models learn from their mistakes, adapting over time. This iterative process ensures that the model's performance consistently improves, delivering more accurate results as it processes more data.

We offer a simple recurring monthly pricing model based on two main elements:
1. A base license fee, chargeable monthly for use of our AI platform2. Tiered monthly charge based on usage (e.g. number of images aggregated over the commitment period with a simple overage fee for any images analysed exceeding the total allowance)

We charge set-up fees for a number of items such as:

1.Model development for additional detection cases beyond the pre-agreed allowance

2.Custom dashboards with personalised KPIs, analytics and triggers

3.Workflow consulting and adaptationAPI systems development support
‍
4.Selection and integration of bespoke edge hardware or software to support image collection

We offer our existing pool of pre-trained detections to all new customers that onboard with Inveniam. For any new customers signing up to 12 months + commitment terms we also develop and deploy new custom models to support a pre-agreed number of use cases on a discounted basis. Get in touch with our Sales team to learn more.

Our platform supports a wide range of visual data, including images, videos and sensor data, as well as meta data. From high-resolution images to streaming videos, we provide comprehensive analysis and predictions for diverse use cases.

We provide a completely transparent and fair pricing model with any unused images in your allowance rolling over to the following month. If you do need to exceed the allowance, we charge a simple overage fee per image. Again, we can customise our pricing model if this doesn’t quite work for you.

We provide the option for you to downgrade or upgrade your usage tier and / or change the length of your commitment period. You can of course cancel the service if you so wish.

We charge additional fees for bespoke items such as creation of bespoke analytics and dashboards, premium SLAs that fall outside of our standard model, or if you require us to rapidly build and deploy custom models for additional detections that fall outside of your annual allowance.

We structure PoC costs to be minimal, and any expenses incurred are offered as credit should you opt for a full-scale deployment with Inveniam. The specific cost varies, influenced by factors such as the quantity and complexity of selected detection cases. More complex cases may necessitate additional effort in areas such as custom model creation, refinement, and evaluation.

The number of use cases we typically incorporate into a Proof of Concept (PoC) varies based on the intricacy of the cases. In most instances, we find that focusing on two to three meticulously chosen detection cases adequately demonstrates the prowess and value proposition of our AI solution. This approach strikes a balance between showcasing capabilities and maintaining clarity.

The process of setting up and delivering a Proof of Concept (PoC) usually takes between 4 to 12 weeks. This variation in time is influenced by factors like project complexity, the number of use cases, and the speed at which input data is collected by our clients.

Onboarding is straight forward. We require a signed copy of our Master Services Agreement, alignment on use cases and detections which to focus our model training and API integration. Onboarding usually takes less than 2 weeks; however, it may take longer if workflow / process adaption is required to support the deployment of the selected use case(s).

We allocate a dedicated Customer Success Manager to every new customer that onboards with us to support with objectives setting, use case exploration, API integration and coordination of resources on both sides to ensure a successful onboard.

The process involves a number of steps: 
‍
1.Explore Challenges and Objectives: We start by understanding the challenges or problems you want to address using computer vision technology. We determine the assets, items, or objects you need to classify and the specific types of detections you're seeking to make. This could include anything from recognising specific objects in images to identifying certain patterns or anomalies.

2.Initial Image Review: We ask you to share some initial images related to your use case. These images are reviewed to assess their suitability for computer vision analysis and to understand the complexity of the task.

3.Image Annotation: Once the initial images are reviewed and deemed suitable, the next step involves collecting and preparing a more extensive dataset of images for annotation. These images are used to train the custom machine learning model. Image annotation involves labelling objects of interest within the images, providing the necessary ground truth data for training the models.

4.Model Training:  We take care of the model training process. Using the annotated dataset, machine learning techniques are employed to train and fine tune a custom model that can accurately classify and detect the objects or patterns you're interested in.

5.Model Deployment:  After the model is trained and reaches a target level of performance, we deploy it into a production environment that supports either real-time or batch processing of the images.It is worth noting that the exact process and time to get a model to production may vary depending on the complexity of the use case, the type of objects being detected, and the level of customisation required for the machine learning model.

We always recommend that you dedicate an internal lead to the computer vision initiative that coordinates resources, image collection and works with internal stakeholders to deploy the solution (such as systems development for API integration).

We utilize the OpenAPI 3.0 standard

Yes, you can use your existing field management applications and tools as we understand the value of integrating with your established workflows.

We offer a multi cloud deployment model as our standard deployment approach, backed by scalable, enterprise grade cloud resources from AWS, Amazon and MS Azure. We can also offer on-premises or edge-based deployments on request based on your operational and use case requirements.

We currently support English and French. Please do get in touch with us if you have a specific language requirement, as we are able to customise our platform on request.

You can either contact your Customer Success Manager directly or log a trouble ticket on the Inveniam service platform.

We provide market leading availability, response, repair time and accuracy SLAs, as well provide a dedicated Customer Success Manager to each account to review and maximise ongoing performance.

Yes, our platform allows you to track status along the data pipeline, including all uploaded and analysed images, results and predictions. Our platform also includes a project monitoring system which informs you on the status of all ongoing development you have requested from our team.