We help you build optical engineering solutions better, faster, and cheaper
Sidekick brings the power of AI to Optical Engineering, enabling users of all skill levels to discover the most optimal solutions for any scenario in minutes.
How Sidekick Works
Sidekick is an intelligent design and sourcing engine for Optical Engineering. Users enter design parameters and output criteria, and Sidekick’s AI engine produces design schematics that hit their performance specifications.
Powerful enough for manufacturing, with a user interface you can learn in minutes.
Users can compare design schematics on a variety of performance metrics, explore trade-offs, and adjust constraints to create alternate scenarios.
Sidekick is the first engineering tool that generates designs for your parameters
INPUT
User specifies parameters e.g. materials, energy sources, vendors
SOLUTION GENERATION
Sidekick’s AI engine assesses all combinations and presents optimal solutions
ANALYZE
Users assess solutions using built-in tools and receive design schematics
Sidekick Use Cases
Scientific Researchers
Rapidly explore a vast number of solutions for any requirement, and expand technical reach with a broad range of built-in skills.
Development Engineers
Conduct in-depth manufacturability analysis and compile detailed tolerancing and design reports.
Application Engineers
Easily assemble compelling documentation demonstrating how your solution meets customer needs and educate customers around design tradeoffs.
SIDEKICK IS POWERED BY FIRST-OF-ITS-KIND AI TECHNOLOGY
Predict
The most common problem in engineering is determining the most advantageous way to solve a problem. Sidekick applies proprietary ML technology to predict the most optimal technical approaches, focusing engineering efforts on solutions with the most benefits.
At Trailhead Analytics, we generate our own training data. Our system maintains revision control and clear traceability for all data and models generated, and our proprietary AI software handles increased complexity with minimal increases in training data or time.
Compute
Sidekick’s ML models greatly enhance the speed and efficiency of our high-fidelity physics code.
Our code is highly parallelized, and can be executed on over 1,000 processors in the cloud to generate solutions in the shortest time possible.
Our system takes care of all of the problem segmentation and parallelization, so there is no need to budget compute times for a problem.
Summarize
At Trailhead Analytics, we pride ourselves on generating actionable information.
After the cloud compute process is complete, we comb through the results and assemble summaries and information to help you make decisions.
Users can also expand summaries by adding new analytics to explore the performance of solution sets in greater detail.
Frequently Asked Questions
Do you need user data to train your machine learning?
No. All of our machine learning models are pre-trained before being deployed to users. Each user’s solutions are completely independent of the requests and solutions of other users.
Is there cross contamination of IP?
Our training doesn’t make use of any existing IP, unless uploaded by authorized vendors. We also don’t make use of patent databases, academic papers, etc. to train our machine learning.
Do you only work with catalog parts?
No. We can generate custom component designs, as well as handle custom services.
Can I request new skills?
Yes - please contact us to request a new skill. One caveat: Sidekick does have a strict training curriculum for the system, so it may take some time to launch this skill.
Can you display volume pricing?
Yes. We can include that in the solution format if it is provided by the vendor.
What information do you provide vendors about users?
We don’t share individual user data with vendors. We make it easy for users to reach out to vendors based on the solutions we provide. Aggregate user data is available to vendors to use in ROI and sales channel analysis.