PATTY TECHNOLOGY INDEX · 2026.09
We publish how AI systems are actually engineered.
From model training through evaluation, inference infrastructure, and applied systems, 21 technical monographs document the decisions, methods, validation, and failure boundaries Patty uses in real engagements.
- Notes
- 21
- Domains
- 04
- Languages
- KR / EN
A capable model is not yet an operable AI system. Training lineage must reach evaluation; evaluated models must run on predictable infrastructure; and inference must enter work through explicit authority and evidence.
- 01Design models and data→
- 02Validate against real work→
- 03Serve under operating constraints→
- 04Connect controlled work
Technical note
Models & Training
- A.01Read note
Pretraining & CPT
A CPT discipline spanning corpus diagnosis, mixing, tokenization, distributed runs, checkpoints, and serving acceptance.
- A.02Read note
SFT Data Pipelines
Design source ingestion, schemas, deduplication, contamination control, human audit, calibration, and immutable exports.
- A.03Read note
RL Alignment: GRPO & RLHF
Design task suitability, rollouts, rewards, GRPO/RLHF choice, reward hacking controls, evaluation, and recovery together.
- A.04Read note
Training Dataset Generation
Design source taxonomy, synthetic generation, independent audit, sensitive-data removal, coverage, and human sampling.
- A.05Read note
Korean-Specialized Modeling
Address morphology, endings, spacing, token efficiency, Korean-English mixing, long sessions, and evaluation gaps together.
- A.06Read note
Custom Model Development
Design the build-versus-adapt decision, organization-specific layers, versioning, safety gates, and deployment boundaries.
Technical note
Evaluation & Benchmarks
- B.01Read note
Benchmark Design & Engineering
Combine representative workloads, scoring rules, contamination controls, human review, and versioning in one evaluation protocol.
- B.02Read note
Model Evaluation & Red Teaming
Manage capability, safety, security, bias, and tool-use failures through threat models and evidence-based release gates.
Technical note
Inference & Infrastructure
- C.01Read note
Distributed GPU Serving & Scheduling
Model request shapes, parallelism, batching, routing, and failure domains together to engineer service levels.
- C.02Read note
KV Cache Architecture
Design block lifecycle, prefix reuse, tiered storage, eviction, and tenant boundaries around the workload.
- C.03Read note
MoE Serving (Wide EP)
Co-design batching, routing, all-to-all communication, and failure scope for wide expert parallel deployments.
- C.04Read note
Multimodal Inference
Engineer resolution, ordering, caching, batching, and output evidence for image, document, and audio inputs.
- C.05Read note
RL Rollout Acceleration
Build a measurable lifecycle across inference workers, sample buffers, reward computation, and trainers.
- C.06Read note
GPU Farm Operations
Unify capacity planning, telemetry, workload isolation, maintenance, and recovery drills in one operating model.
Technical note
Applied AI
- D.01Read note
Korean-Specialized RAG
Design the path from Korean document ingestion through structure-preserving chunking, hybrid retrieval, reranking, evidence, and access control.
- D.02Read note
Document Intelligence Pipelines
Preserve layouts, tables, forms, and signatures across HWP, HWPX, PDF, and images while linking extraction, validation, and approval.
- D.03Read note
LLM-Native Browser
Engineer DOM, accessibility, and visual state with session isolation, tool authority, approvals, and reproducible execution records.
- D.04Read note
Agent Orchestration
Design task decomposition, tool contracts, state, approvals, retries, and compensating transactions as an observable graph.
- D.05Read note
AI Workforce Transformation
Connect work decomposition, risk tiers, roles, approvals, quality measurement, and change management in an operating model.
- D.06Read note
ViewSpec Rendering
Place typed component vocabulary, data binding, accessibility rules, policy validation, and visual regression at the rendering boundary.
- D.07Read note
On-Premises Deployment
Design network boundaries, supply chain, secrets, model import, observability, patching, backup, and handover from a threat model.
Begin with the bottleneck in your current system.
We map representative workloads, system constraints, and acceptance evidence before proposing an intervention.