apidays Munich 2025 - Agentic AI: A Friend or Foe?, Merja Kajava (Aavista Oy)
Download as PPTX, PDF0 likes50 views
Agentic AI: A Friend or Foe? Merja Kajava, Founder at Aavista Oy apidays Munich 2025 - Accelerate AI Use Cases with APIs July 2 & 3, 2025 ------ Check out our conferences at https://www.apidays.global/ Do you want to sponsor or talk at one of our conferences? https://apidays.typeform.com/to/ILJeAaV8 Learn more on APIscene, the global media made by the community for the community: https://www.apiscene.io Explore the API ecosystem with the API Landscape: https://apilandscape.apiscene.io/
apidays Munich 2025 - Agentic AI: A Friend or Foe?, Merja Kajava (Aavista Oy)
- 1. Aavista Oy Merja Kajava apidays Munich 2025 Agentic AI – A Friend or a Foe
- 2. “What is the carbon footprint of a ChatGPT query compared to a Google query?” Photo credit: NASA / Alex Gerst
- 3. Vanderbauwhede (2024): “AI-generated answers to conventional search queries dramatically increase the energy consumption. By our estimates, energy demand increase by 60-70 times.” Vanderbauwhede, W. (2024). Estimating the Increase in Emissions caused by AI-augmented Search. https://doi.org/10.48550/ARXIV.2407.16894
- 4. Moving from chatbots to Agentic AI
- 5. Tools Orchestration Model Integrate to storage and 3rd party services Reason with goals Goals, profiles and guidance What are the elements of Agentic AI
- 7. Agent-to-Agent (A2A) by Google Model Context Protocol (MCP) by Anthropic “UDDI for Agents” for agent discovery “BPEL for Agents” for orchestration NLWeb by Microsoft Agent Agent Agent <MCP> Tool <MCP> Tool <MCP> Tool RSS Schema.org Protocols are emerging for agent connectivity MCP MCP MCP A2A A2A MCP API MCP Servers <NLWeb> MCP Servers Agent Communication Protocol (ACP) by IBM Agent Connect Protocol (ACP) By Agntcy Collective
- 8. Sustainability AI governance Security and privacy Testing Performance Data quality Cost
- 9. Nothing exists until it is measured Niels Bohr
- 11. Calculate the carbon footprint Cloud providers Data centers Service providers IPaaS API SaaS AI AWS Google Cloud Microsoft Multi-cloud Applications Carbon footprint calculators by cloud providers. Multi-cloud calculators. Open-source calculators, for example Code Carbon. Commercial calculators, for example from Dynatrace and Splunk.
- 12. “Processing a single query might involve over 50 LLM API calls. The calls tend to be very narrow and specific. Different kinds of calls may be to different models.” NLWeb – Chat query example https://github.com/microsoft/NLWeb/blob/main/docs/life-of-a-chat-query.md Case NLWeb – Chat query example Sustainability Performance Cost
- 14. AI Energy Score - Extractive Q & A by Huggingface Each Gen AI model is different Luccione et al. (2024): “Multi-purpose models are more energy-intensive.” Luccioni, S., Jernite, Y., & Strubell, E. (2024). Power Hungry Processing: Watts Driving the Cost of AI Deployment? The 2024 ACM Conference on Fairness, Accountability, and Transparency, 85–99. https://doi.org/10.1145/3630106.3658542
- 15. Multimodal AI brings new challenges Multimodal streaming “Agent architecture with bi-directional event streaming running 24/7”
- 16. Multimodal AI From text prompt to images,videos and speech Luccioni, S., Jernite, Y., & Strubell, E. (2024). Power Hungry Processing: Watts Driving the Cost of AI Deployment? The 2024 ACM Conference on Fairness, Accountability, and Transparency, 85–99. https://doi.org/10.1145/3630106.3658542 “Tasks involving images are more energy- and carbon-intensive compared to those involving text alone.”
- 17. In search of energy certificate for AI agents Photo credit: Energiakartoitus Oy
- 18. Key takeaways Choose the minimal solution - YAGNI Follow up the costs Measure the carbon footprint
- 19. Technology is the answer, but what was the question? Cedric Price
- 20. The Data Refinery Company Merja Kajava https://www.linkedin.com/in/merjakajava
Editor's Notes
- #2: Photo credit: NASA / Alex Gerst
- #3: The paper estimates the energy demand of queries to the BLOOM model, a 176B parameter model, and OpenAI's GPT-3, which is of similar complexity.
- #5: An AI agent is autonomous software that can perceive, reason and act.
- #7: MCP by Antropic – Introduced November 2024 - JSON-RPC Client-Server A2A Agent-to-Agent protocol (by Google) - Introduced April 9, 2025 NLWeb by Microsoft - Introduced May 19, 2025
- #11: Carbon footprint calculators as part of the cloud billing. Code Carbon (open-source) https://codecarbon.io Calculators from Dynatrace, Splunk. Cloud Computing Carbon Emissions by Climatiq (multi-cloud). Open-source multi-cloud - Cloud Carbon Footprint
- #19: Are agents always the solution?