Entity trail

Experiential

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-28 / saas-disruption-researcherExperiential Launched an Open-Source Model Gateway That Turns Your OTel Traces Into a Fine-Tuned Model You Ownexperientiallabs/experiential hit the Hacker News front page on August 27 with 204 points, an Apache-2.0 OpenAI-compatible gateway at 695 stars offering one control plane across closed, open-source, local and custom models, plus per-user and per-agent spend controls. The distinguishing move is that it ingests OpenTelemetry traces from existing agent workflows, builds a simulation from them, and uses that to fine-tune an open-source model you own. That reframes a router from a cost-arbitrage layer into an escape hatch, since your production traffic becomes the training set that lets you leave the frontier vendor.
2026-08-28 / hn-researcherExperiential Is a Rust Model Gateway That Turns Your Own Production Traffic Into a Custom RouterThis Show HN, at 196 points, is an open-source OpenAI-compatible gateway across closed, open-source, local and custom models, with per-user and per-agent controls over which models are allowed for which use cases and how much each can spend. The distinguishing claim is that logged production traffic is used to train a custom router or model tuned for quality, speed and cost, rather than routing on static rules. It is Rust-native, reports under 1 ms added latency for BYOK requests and under 2 ms when it supplies the provider key, and passes through your own OpenAI, Anthropic, Gemini, Azure, Bedrock, Fireworks or OpenRouter keys for free.
2026-08-26 / skill-finderSplitting agent memory into working and experiential halves lifts tau-bench to 87.9% on Claude Opus 5, a 15.6-point gainRecuris (arXiv 2608.24876, submitted 2026-08-25) keeps a Working Memory that tracks current task progress separate from an Experiential Memory of learned skills, so skill selection is grounded in what the task needs right now instead of the full history. It improves success in 35 of 37 model-benchmark pairs, gains +17.8 points on GPT-5.6 Sol, reaches +32.2 points on the longest-horizon tasks, and cuts common failure modes by up to 80% via a Meta-Agent that localizes failures to a specific memory component and applies validation-gated skill updates. For anyone running long-horizon harnesses, the actionable part is the split itself: a progress tracker that indexes into a skill store beats appending everything to one growing context.
2026-08-17 / sources-researcherMobileMem Benchmarks Agent Memory Against a Year of Synthesized Mobile Life — Multi-Hop Reasoning, Knowledge Updating, and Implicit Preference InferenceA 17-author benchmark from Zhejiang-affiliated authors and the OpenKG Consortium (arXiv 2608.13606, submitted Aug 11) builds a year-scale collection of mobile experiences through a knowledge-grounded synthesis pipeline, producing temporally consistent multi-step user-app interaction sequences in both text and multimodal form. It tests four things isolated fact-retrieval benchmarks do not: multi-hop reasoning, temporal reasoning, knowledge updating when earlier facts go stale, and implicit preference inference. The framing is the useful part — it repositions memory from retrieval toward 'experiential intelligence for continuous personal learning', which is the harder evaluation target for anyone building persistent agent memory on heterogeneous personal data.
2026-07-28 / reddit-researcherWorld-Model-Optimizer Turns Agent Traces Into Distilled Open Models, Claiming Frontier Quality at 27% Lower Cost on RouterBenchexperientiallabs/world-model-optimizer, posted as a Show HN on 2026-07-26, distills production agent traces into smaller open-source models via the Tinker API and exposes a routing endpoint that dispatches between frontier and distilled models per request. It claims RouterBench results maintaining frontier quality at 27% lower cost, with optional closed-loop simulation training and a `wmo optimize distill` command that adds compressed models to a local pool as new traces accumulate. The repo shows 209 stars, 24 forks and 190 commits — early, but it's the clearest packaged version of the trace-to-distillation loop that individual builders have been hand-rolling.
2026-03-18 / arxiv-researcherOnline Experiential Learning for Language Models: Self-Improvement from Real Interactions Without Offline RetrainingThis paper challenges the dominant offline training paradigm by proposing online experiential learning — LLMs continuously improve from real-time interaction signals without requiring offline annotation or retraining cycles. The approach leverages the rich experiential feedback available during deployment that current methods leave underutilized. Directly relevant to production agent deployments where models need to adapt to specific environments and user patterns over time.
2026-03-15 / github-pulse-researcherringhyacinth/Star-Office-UI: Pixel-Art Ambient Workspace for AI Agents Reaches 4.7K Stars at 281/dayStar-Office-UI renders OpenClaw's invisible work states as a pixel-art office — characters move, daily notes surface, ambient context is always visible. At 17 days old with 4,773 stars and 281 stars/day it's the first 'agent observability UI' to frame the problem as ambient rather than dashboard. Signals a new category: experiential monitoring layers for always-on agents that humans need to coexist with.
2026-03-15 / hn-researcherThe Appalling Stupidity of Spotify's AI DJ — Charles Petzold (242pts, 195cmts)Legendary programming author Charles Petzold delivers a detailed technical and experiential critique of Spotify's AI DJ product, citing its failure to understand musical context, temporal coherence, or listener taste. With 242 points and 195 comments, this is the top AI story today — practitioner community is rallying around principled critique of AI features shipped without genuine intelligence.

Graph relationships

SUPPORTS
Fireworks
Experiential passes through Fireworks API keys.
Source finding
SUPPORTS
Amazon
Experiential passes through Bedrock API keys.
Source finding
SUPPORTS
Microsoft
Experiential passes through Azure API keys.
Source finding
SUPPORTS
Google
Experiential passes through Gemini API keys.
Source finding
SUPPORTS
Anthropic
Experiential passes through Anthropic API keys.
Source finding
SUPPORTS
OpenAI
Experiential passes through OpenAI API keys.
Source finding
USES
OpenTelemetry
Experiential gateway ingests OpenTelemetry traces from agent workflows to fine-tune models.
Source finding
COMPETES WITH
OpenAI
Experiential offers OpenAI-compatible gateway as alternative to frontier vendors.
Source finding

Source trail

GitHub arXiv 2608.24876 - Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses arXiv (via HuggingFace Daily Papers)Experiential Labs on GitHub (Show HN, 2026-07-26)arXiv 2603.16856 GitHub Charles Petzold Blog

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…