Electric Brain - Your AI laboratory for hire!

At Electric Brain, we specialize in delivering custom AI technology on fixed time-lines and fixed budgets. Whether you want to develop a prototype of new AI technology, automate a process at your business, or simply improve accuracy on a model you already have, we have talented, deep-learning experts who are ready to deliver new technology for you. Electric Brain has the experience and industry leading tools to deliver real solutions that drive real business value… without taking years to do it.

Transparent Pricing

We make it as easy as possible to predict pricing for your AI project even before you reach out to us. All projects will require an AI research phase. Depending on your team composition and requirements, you will usually want an integration phase as well. Many projects also require a data collection phase before the project begins. Remember these numbers are meant to provide guidance only – please contact us for an exact quote!

SERVICES
DESCRIPTION
ESTIMATED PRICE RANGE

Small AI Research Project

If your data can be put into a spreadsheet (e.g. numbers, categories, dates, and text)

$7,500 to $15,000

Medium AI Research Project

If your project involves images or audio, or unsupervised learning (clustering, segmentation, similarity/difference)

$15,000 to $25,000

Large AI Research Project

If your project involves videos, very big data (+1 billion rows), or is a known hard problem (like chat-bot engines)

$25,000 to $50,000+

Small AI Integration

If you need an integration with an existing web-application with few required modifications, or your team is doing most of the work

$5,000 to $10,000

Medium AI Integration

If your model needs to run on a mobile device, or you require a prototype built from scratch to surround your AI technology

$15,000 to $30,000

Large AI Integration

If you need integration in a complex enterprise environment with multiple systems and platforms

Please contact

Small Data Annotator

If you need a simple tool for humans to ‘train’ your AI by enhancing data you already have

$10,000 to $15,000

Medium Data Annotator

If you need a tool that can simultaneously ‘train’ the AI while operating the business process in a production environment

$15,000 to $35,000

Large Data Annotator

If you need a tool to ‘train’ the AI and it must be deployed in a complex enterprise environment with many involved stakeholders

Please contact

AI Research Process

STEP 1: Brainstorm and discuss your AI ideas

It all starts with a great idea. Maybe you have a process you want to automate. Maybe you have some data that you want to monetize. Or maybe you just have something never before seen that you want to invent. Any way it is, our team will sit down with you and flesh out which ideas are feasible, and how much data is likely required to train a model.

STEP 2: Create an official proposal

Once you have picked a artificial intelligence project that you want to proceed with, Electric Brain will create a proposal containing price, time-line, and technology details for the project.

STEP 3: Collect the data

Now the actual project begins. While you may already have all the necessary data for your machine learning project, we have found that the most promising, innovative artificial intelligence projects usually require people to manually collect or enhance data. This is done in two steps. First, we create a small tool known as an annotator, which is used for building the dataset quickly and efficiently. Second, we hire lots of cheap data-entry people to collect the data.

STEP 4: Train the neural network

With data in hand, the Electric Brain team can now get to work on your neural network. We will hand-pick and test potential architectures for solving your problem. When a promising neural network is found, we will then hand-tune it over the course of a few weeks. This ensures that you get the highest accuracy model for your problem.

STEP 5: Validate the neural network

Nearing completion, Electric Brain now validates the neural network model by testing it with further data. We prepare a report, indicating how the neural network performance is impacted by different variables, such as the size of the dataset or the processing power. From this report, we will be in a position to tell you what is required to achieve a given level of performance, whether it’s more data, more processing power, or both.

STEP 6: Integrate the solution

With a working neural network model in hand, you can now integrate it into a final solution. We provide you with a small RESTful API server known as a ‘bundle’, which does your core machine learning. This makes it super easy for your programmers to integrate it into a larger solution, no matter what stack you use.

STEP 7: Monitor the performance

The most important step in any artificial intelligence solution is to monitor the performance in a production environment. This ensures that your model is performing well on new data. Sometimes, additional tuning may be required if the model performs differently in production then it did during testing.

STEP 8: Celebrate the launch of your artificial intelligence solution

At the end of the project, we can all gather around to a bottle of Champagne to celebrate the launch of your new AI!

LET'S CHAT