Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 7 hours
Course Outline
Transitioning from Generative Text to Structured Decisioning
- Output specifications of decision models
- Distinguishing heuristic decisioning from reasoning LLMs
- Latency and computational cost of constrained decisioning
The Ecosystem of Open-Weight Decision Models
- Strands Decider 2B released by AWS Labs
- Overview of JEV and GEV open-weight model families
- Pplx-Decider, Liquid D1, and Clef architectures
- Analysis of licensing terms and deployment considerations
Architectural Components of Decision Models
- Utilization of the LLM backbone without a language generation head
- Scoring mechanisms via pointer heads over candidate options
- Derivation and interpretation of confidence metrics
Implementation of Initial Decision Processes
- Local environment configuration on CPU and consumer-grade GPUs
- Formulating inquiry structures and option sets
- Batching multiple queries in a single processing pass
- Establishing decision thresholds and defining fallback protocols
Evaluation of Decision Quality and Reliability
- JevBench and other decision-specific benchmarking suites
- Balancing accuracy against statistical calibration
- Brier score analysis and confidence reliability assessment
- Identifying operational tasks where decision models are unsuitable
Requirements
- Foundational knowledge of large language models and prompt-based application development.
- Proficiency with command-line interfaces and Python scripting.
- No prior machine learning training experience is required.
Target Audience
- AI engineers evaluating decision models for agent-based government workflows
- Software developers seeking to replace LLM calls with bounded decision logic
- Solution and platform architects designing robust agent architectures
- Technical product managers assessing decision model options for public sector deployment
