The challenges associated with getting AI models into production are well known. What developers need now are ways to address those challenges. Ramtin Davanlou, from Accenture*, presents his thoughts around how developers can get their models in production and do it more quickly during his presentation at the oneAPI DevSummit for AI 2022.
Davanlou first discusses the development cycle to deploy AI systems and its associated challenges. He emphasizes that to move through this cycle successfully requires the combination of both hardware acceleration and the software optimizations provided by oneAPI.
He then talks about how Intel has created AI reference kits, in conjunction with Accenture*, to connect developers with the Intel software technology to solve problems in key business domains. The main objective of the kits is to reduce the time required to deploy AI models.
Each kit includes the following:
- Solution Brief: An overview of the value proposition, describing the problem, solution, and impact.
- Developer Guide: Recommendations for frameworks, algorithms, data processing techniques, hyper -tuning, quantization, deployment, including showing how to build the ML pipeline.
- Code Repository: GitHub code snippets, configurations, datasets, and libraries before and after optimization.
- Platform Architecture: A guide for setting up the best performing compute architecture.
- Benchmarking Results: Performance gains & metrics showing impact of oneAPI optimizations.
Davanlou concludes by discussing these four industry use cases and their AI reference kits, which are available now.
- Predictive Asset Maintenance: Predict the probability of failure and proactively maintain assets to avoid outages, downtimes and operational costs.
- Visual Quality Inspection: Use computer vision to detect defects and reduce quality inspection time and costs.
- Customer Care Agent Intent Enablement: Enable virtual agents to understand user intents in automated conversations using Natural Language Understanding (NLU).
- Intelligent Document Indexing: Reduce human capital costs and manual intervention for classifying massive volumes of incoming documents ingested into the organization.
Watch the presentation below to learn how to accelerate AI development cycles by using the tools and framework optimizations that are part of Intel’s AI software portfolio.
See the video: Enabling AI Developers on Their Journey to Scale with Intel AI
All models are optimized with Intel AI Tools powered by oneAPI for faster training and inferencing performance using less compute resources. The AI reference kits use components from Intel's AI software portfolio, including Intel® oneAPI AI Analytics Toolkit and the Intel® Distribution of OpenVINO™ toolkit.
To learn more about the AI reference kits, visit https://www.intel.com/content/www/us/en/developer/topic-technology/artificial-intelligence/reference-kit.html
About our expert
As a part of Accenture’s Applied Intelligence group, Ramtin Davanlou leads the development of global offering and solutions in analytics and artificial intelligence. His key responsibility is to drive growth in AI through innovation and thought leadership. He works with ecosystem partners and provides optimized solutions and technology architecture recommendations to clients in various industries.
Susan is a Product Marketing Manager for AIML at Intel. She has her Ph.D. in Human Factors and Ergonomics, having used analytics to quantify and compare mental models of how humans learn complex operations. Throughout her well-rounded career, she has held roles in user-centered design, product management, customer insights, consulting, and operational risk.