Resources # Whitepapers & benchmarks Source: Title: Resources | Goodmem > Open, web-native technical whitepapers and benchmarks from the team building GoodMem — the memory layer for agentic AI. Free to read, no sign-up. Engineering reports from the team building GoodMem — benchmarks and results on real workloads. Free to read, no sign-up. [GoodMem · TCO Brief](https://goodmem.ai/resources/agent-memory-tco) ## [The Total Cost of Agent Memory](https://goodmem.ai/resources/agent-memory-tco) [An itemized account of GoodMem's on-disk footprint, priced against measured token savings at current list prices — with an interactive calculator for your own workload.](https://goodmem.ai/resources/agent-memory-tco) [**saved per conversation**](https://goodmem.ai/resources/agent-memory-tco) [$1.18](https://goodmem.ai/resources/agent-memory-tco) [**stores a document's footprint**](https://goodmem.ai/resources/agent-memory-tco) [13¢ / yr](https://goodmem.ai/resources/agent-memory-tco) [**savings to infrastructure**](https://goodmem.ai/resources/agent-memory-tco) [9:1](https://goodmem.ai/resources/agent-memory-tco) [August 2026 Read the report](https://goodmem.ai/resources/agent-memory-tco) [GoodMem · Benchmark](https://goodmem.ai/resources/cutting-agent-token-burn) ## [Memory That Pays for Itself](https://goodmem.ai/resources/cutting-agent-token-burn) [Controlled A/B benchmark — GoodMem's memory layer cut aggregate token use 28% and reasoning iterations 23%, at no measurable loss of answer quality.](https://goodmem.ai/resources/cutting-agent-token-burn) [**token burn**](https://goodmem.ai/resources/cutting-agent-token-burn) [−28%](https://goodmem.ai/resources/cutting-agent-token-burn) [**reasoning steps**](https://goodmem.ai/resources/cutting-agent-token-burn) [−23%](https://goodmem.ai/resources/cutting-agent-token-burn) [**quality change**](https://goodmem.ai/resources/cutting-agent-token-burn) [≈ 0](https://goodmem.ai/resources/cutting-agent-token-burn) [June 2026 Read the report](https://goodmem.ai/resources/cutting-agent-token-burn) [GoodMem · Technical Brief](https://goodmem.ai/resources/ingestion-quality) ## [Garbage In, Garbage Out](https://goodmem.ai/resources/ingestion-quality) [Ingestion sets the ceiling on RAG quality. On-prem OCR that never leaves your network, and page-image capture across PDF and all three Office formats — Excel included.](https://goodmem.ai/resources/ingestion-quality) [**page images**](https://goodmem.ai/resources/ingestion-quality) [4 formats](https://goodmem.ai/resources/ingestion-quality) [**document OCR**](https://goodmem.ai/resources/ingestion-quality) [On-prem](https://goodmem.ai/resources/ingestion-quality) [**page-faithful**](https://goodmem.ai/resources/ingestion-quality) [Excel](https://goodmem.ai/resources/ingestion-quality) [June 2026 Read the report](https://goodmem.ai/resources/ingestion-quality) [GoodMem · Security & Trust Brief](https://goodmem.ai/resources/enterprise-infrastructure) ## [Built Like Enterprise Infrastructure](https://goodmem.ai/resources/enterprise-infrastructure) [Resource-aware access control, scoped API keys, and a verifiable supply chain on Java and PostgreSQL. Built by ISO 27001-certified PAIR Systems, with its SOC 2 Type II audit in progress.](https://goodmem.ai/resources/enterprise-infrastructure) [**PAIR Systems certified**](https://goodmem.ai/resources/enterprise-infrastructure) [ISO 27001](https://goodmem.ai/resources/enterprise-infrastructure) [**provenance**](https://goodmem.ai/resources/enterprise-infrastructure) [SLSA L3](https://goodmem.ai/resources/enterprise-infrastructure) [**one data layer**](https://goodmem.ai/resources/enterprise-infrastructure) [PostgreSQL](https://goodmem.ai/resources/enterprise-infrastructure) [September 2026 Read the report](https://goodmem.ai/resources/enterprise-infrastructure) ## Frequently asked questions Direct answers on what GoodMem stores, how retrieval and access control work, deployment and operations, pricing, and the things it deliberately does not do. [Read the FAQ](https://goodmem.ai/resources/faq) Interactive technical brief ## The Architecture of Agent Optimization An interactive map of the agent stack: what each layer contributes to quality and cost, and the measured evidence for engineering it. [Open the full brief](https://goodmem.ai/resources/agent-optimization-architecture) Explore the agent architecture **Layer 03 · Agent component LLM** PAIR trained a 35B-class MoE that outscored GLM‑5.2 on a controlled BI-agent benchmark. It reached 91% of Opus 4.8’s measured quality in medium-thinking mode at 35× lower serving cost. [View model benchmark](https://goodmem.ai/resources/agent-optimization-architecture#llm) **Layer 03 · Agent component Harness** [Zhang et al.](https://arxiv.org/abs/2605.23950) report 7.8× more performance variance from harness configuration than model choice. PAIR engineers custom harnesses for GoodMem customers as a service. **Layer 02 · Retrieval models Rerankers** GoodMem’s Fine-Tuning automatically derives optimized embedding and reranking models from two inputs: memory content and agents’ access patterns. [View Fine-Tuning study](https://goodmem.ai/resources/agent-optimization-architecture#retrieval-models) **Layer 02 · Retrieval models Embedders** GoodMem’s Pipeline Optimization automatically selects the best combination of embedders and rerankers for each agent’s unique context. **Layer 02 GoodMem Context / Memory** Controlled tests cut aggregate token burn 28% and agent steps 23%, with no measurable change in quality. Heavy retrieval tasks saved 30–66%. [View memory study](https://goodmem.ai/resources/agent-optimization-architecture#goodmem) **Layer 01 PostgreSQL Neon · AlloyDB** Battle-tested reliability, security, and industry support. An open data layer with hundreds of deployment paths. Each colored block links to its evidence in the full brief. [Explore the complete architecture](https://goodmem.ai/resources/agent-optimization-architecture)