HeyJaredIris ai
AI Harnessing for expert knowledge. For industries where a wrong answer isn't an option. Iris.ai is a Norwegian AI company, founded in Oslo in 2015 as the country's first AI startup, with 11 years of research behind the product.
- Indexed videos, last 90 days
- 4
- Latest publication
- Sep 15, 2026
- Audience
- ~1.1K subscribers
- Earliest in this view
- Aug 24, 2026
Latest videos
When it comes to enterprise AI, there are two engineering philosophies... (opens the original)
Read excerpt
When it comes to enterprise AI, there are two engineering philosophies: let it break and fix it later, or build it with strict guardrails from day one. In regulated industries, the first option is a compliance disaster. In the latest segment of our AWS conversation alongside Matthew Thomson, @Steven Fung asked a critical question: What does the EU AI Act actually mean for technical architecture, auditability, and traceability? Victor Botev’s answer highlights the exact engineering philosophy we
Are we repeating the biggest mistake of the Big Data era with generative AI? (opens the original)
Read excerpt
In the latest segment of our AWS conversation alongside Matthew Thomson and Victor Botev, Steven Fung highlights a critical parallel: Big Data gave us modern analytics, but it often lacked the expert knowledge that businesses could actually trust. Today, enterprise AI is hitting the exact same wall. While LLMs offer incredible flexibility, organizations are missing the control layer. For AI to work in production, domain experts must have absolute governance over what feeds into the models and ho
We built a "lie detector" for enterprise AI. (opens the original)
Read excerpt
Here is the reality of deploying generative models: you can explicitly instruct an AI to only use the enterprise context you provided, and it will still hallucinate by pulling from its own parametric knowledge. In a sandbox, that is an annoyance. In a regulated industry, that is a compliance failure. In the latest segment of our AWS conversation alongside Matthew Thomson and Steven, Victor Botev breaks down the missing control piece in enterprise AI workflows. Since 2015, our team has been devel
The infrastructure is ready. The models are powerful. So why is enterprise AI still hallucinating? (opens the original)
Read excerpt
That is the exact question Matthew Thomson from AWS addressed in our recent conversation with Steven Fung and Victor Botev. The reality? Enterprise data layers simply were not built for AI. When enterprise AI projects fail, the compute and the models are rarely the bottleneck. The root cause is almost always an unready data layer. Unstructured knowledge, research papers, regulatory filings, technical manuals, and legacy databases, is just too fragmented for AI to reason over reliably. To move fr
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
~1.1K subscribers
Measured Sep 19, 2026
Source's subscribers, not the number who saw an individual piece.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.