Hardware failures interrupted training runs. Automated recovery cut model training time by 40%.
A generative AI startup serving 10 million monthly active users with a conversational answer engine that synthesizes web data into natural language responses.
As model complexity grew, developers spent excessive time managing infrastructure and recovering from hardware failures during long training runs....
“Using large language models, we can capture human language understanding and reasoning capabilities into one model. That, combined with the facts on the internet, has helped us build our answer engine. Essentially, we orchestrated a traditional search index (facts engine) and a reasoning engine (large language model) together to build the world’s first conversational answer engine.”
Perplexity is a technology company that specializes in artificial intelligence and machine learning solutions.
Cloud computing platform and on-demand infrastructure services.
Related implementations across industries and use cases
Manual fixes for node failures held back training. Automated recovery now drives 4x faster customer deployments.
Researchers were stuck maintaining hardware. Moving to managed infrastructure freed them to build and cut training costs by 90%.
Redacting data for security broke reasoning. Now, private channels allow full context, enabling natural dialogue over rigid menus.
Engineers manually correlated alerts across systems. AI agents now diagnose issues and suggest fixes, cutting recovery time by 35%.
Minor edits required days of crew coordination. Now, staff use avatars to modify dialogue and translate languages instantly.
Lab supply orders were handwritten in notebooks. Digital ordering now takes seconds, saving 30,000 hours for research annually.
Experts spent 15 minutes pulling data from scattered systems. Natural language prompts now generate detailed reports instantly.