Offloading bounded decisions cut agent LLM calls by 29.6% on 360 SOP-Bench cases

Offloading bounded decisions to a decision model cut LLM calls by 29.6% across 360 SOP-Bench cases spanning referral abuse detection, email intent, and patient intake, according to an experiment run by Proceda creator Qwen 3.8 27B with hosted Jev. Completion checking alone saved 23.8% of LLM calls but finished 356/360 tasks correctly versus 357/360 for the baseline, and including final structured-plan preparation cost lowered the estimated cost saving to 24.6%. The author states the combined result comes from this 360-case subset and does not establish a general accuracy improvement.

Disclosure: I built Proceda, the SOP-execution harness used here, and ran these experiments. I wanted to see how much work an agent could hand to a decision model when the answer is narrower than general-purpose text generation. I tested Qwen 3.8 27B with hosted Jev on 360 SOP-Bench cases across referral abuse detection, email intent, and patient intake. I tried two insertion points: With both changes enabled: There is a tradeoff worth separating out: completion checking alone saved 23.8% of LLM calls but finished 356/360 tasks correctly, versus 357/360 for the baseline. The combined result above is from this particular 360-case subset; it does not establish a general accuracy improvement. Including the final structured-plan preparation cost also reduces the estimated cost saving to 24.6%. The useful architectural pattern is to supply evidence, define allowed choices, validate the result, and keep an LLM fallback. The implementation and experiment reports are public: Which decisions in your agents have you successfully offloaded, and how do you validate confidence thresholds before allowing them to bypass the main LLM?