◆ Gentoo Logic · Modeling Warehouse

Architecture comparison

Cognition (Devin) vs Poolside

Poolside runs closer to the foundational model (Native (own / self-hosted weights)) than Cognition (Devin) (Direct API). No shared foundational model — different bets.

Cognition (Devin)

AI-coding

Proximity
Direct API
Models
Anthropic · ClaudeOpenAI · GPT
Cloud · datastore
AWSPostgreSQLvector
Compliance
SOC 2
Architecture

Autonomous SWE agent (Devin) orchestrating Claude + GPT over its own long-horizon planning and code-execution sandboxes.

Poolside

AI-coding

Proximity
Native (own / self-hosted weights)
Models
proprietary / self-built models
Cloud · datastore
AWSApache Iceberg (central data/experiment layer)Object storage over Iceberg (likely S3-compatible, inferred from AWS + Iceberg usage)Kubernetes-native config/state stores (e.g., etcd; inferred from 'stacked under a single Kubernetes orchestrator')
Compliance
Enterprise-grade, in-VPC / on-prem isolation (docs + platform marketing)Air‑gapped deployment support (AWS re:Invent talk)Customer-owned VPC / data residency controls (docs)No explicit public claims of SOC 2 / ISO 27001 / HIPAA in available materials
Architecture

Poolside builds and trains proprietary foundation models via its internal **Model Factory** (Titan training stack on a ~10K GPU cluster, Apache Iceberg data layer, Kubernetes orchestration) and then deploys those models and agentic systems fully inside customer VPCs/on‑prem through the Poolside Platform, with additional distribution via AWS Bedrock and Trainium-backed inference.[1][2][7][8][9][5][6][11] The production architecture is therefore a proprietary, Kubernetes-based ML/agent stack running on AWS (and customer infrastructure) with an Iceberg-centered data plane and tightly integrated model training, evaluation, and deployment pipelines.[1][2][7][8][9][11]

← Full orbit map · Score your own stack →