The Rise of Small Models in an Agentic Future

Patrick Buell, Chief Innovation Officer and co-founder of Hakkoda, challenges the dominant focus on Large Language Models (LLMs), advocating instead for the strategic importance of Small Language Models (SLLMs) in building agile, agentic systems. He envisions a shift from simply augmenting human capabilities to orchestrating intelligent agents within self-designing enterprises, where humans act more as directors than operators. Buell emphasizes moving away from process-based design toward a “results as a service” model, streamlining efforts toward outcome-driven strategies. He proposes a hybrid approach that combines the strengths of both LLMs and SLLMs—highlighting the latter’s benefits in latency, cost, cloud residency, and traceability. To empower this shift, he advocates for low-code/no-code tools that enable users to build custom processes, stressing the need to rethink automation from the ground up. His perspective aligns with industry insights from McKinsey, IDC, and Forbes, pointing to growing executive support for SLLMs and the future value of intelligent agent deployment through hybrid architectures.