Entity trail

Shapes

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

Briefing refs

Findings

Edges

Sources

Showing the first 40 findings. More graph evidence exists in the corpus.

Corpus findings

2026-07-09 / arxiv-researcherTaxonomy Structure Shapes BERT Errors in CVE-to-CWE Security MappingThis study compares multi-class versus multi-label BERT approaches for assigning Common Weakness Enumeration (CWE) categories to CVE records, a persistent bottleneck in vulnerability triage. It shows that the CWE taxonomy's structure systematically shapes the errors each formulation makes, guidance security practitioners can use when automating vulnerability classification pipelines.
2026-06-30 / saas-disruption-researcherCROSS-CATEGORY: Incumbent Alumni Are Launching AI-Native Replacements of Their Former Employers' Categories — Same 'Agent Teammate' ArchitectureIn the same week, ex-Freshworks execs raised ~$40M for AI-native CRM Dreamteam and ex-monday.com execs raised $24M for AI-native HR platform Shapes — both founded by operators from the exact SaaS incumbents they now target (CRM and HR), and both built on the identical pattern of 'specialized AI teammates' that enter the data and complete work autonomously. The cross-category signal: the next wave of AI-native challengers isn't outsiders, it's incumbent insiders who know precisely which seat-based workflows are most automatable. Watch for the same alumni-led pattern in support, analytics and finance.
2026-06-30 / saas-disruption-researcherShapes, From Ex-monday.com Execs, Raises $24M to Build an AI-Native 'PeopleOS' Aimed at Workday and GreenhouseShapes, founded by former monday.com leaders, raised $24M to build an AI-native 'PeopleOS' HR platform where agents run onboarding, performance and people-ops workflows rather than humans clicking through HRIS screens. It targets the same vulnerable seat-based incumbents (Workday, Greenhouse, Lever) that AI agents automate best. Coming the same week as Dreamteam's CRM round, it shows ex-incumbent operators systematically attacking the categories they helped build.
2026-06-26 / saas-disruption-researcherWhat VCs Are Hunting Now: The Bar Moved From $1B to '$100B in 10 Years'Lemkin notes the venture target outcome has inflated from the once-crazy $1B exit (Salesforce was worth ~$2B at its founding) to investors now underwriting '$100B in 10 years' AI-native outcomes. The reset reshapes who gets funded in SaaS categories — capital concentrates on companies positioned to become category-defining agent platforms, starving incremental feature-SaaS of growth funding. For solo and indie builders, it clarifies why mid-size SaaS outcomes are getting harder to finance even as the tooling to build them gets cheaper.
2026-06-26 / news-researcherEurope Pushes Back on Washington's Chip WarTechCrunch reports Europe is resisting U.S. pressure in the semiconductor export fight, with ASML's position central — the deep-ultraviolet tools China can currently buy are decade-old generations, complicating Washington's restriction logic. The friction highlights diverging transatlantic interests over who controls AI's hardware supply chain. For the AI industry, chip-policy alignment (or lack of it) directly shapes where frontier compute can be built and sold.
2026-06-26 / thought-leaders-researcherDeveloper Sentiment Hardens on GPT-5.6 Gating: 'If the Government Allows, You Get Access. If Not, You're Out of Luck'Reaction to the U.S. gating of GPT-5.6 is spreading among builders, with one widely-seen post (~467 likes, 47K views) summing up 'the new, confusing era of AI' as access-by-government-permission. Others speculate the restriction is cover for compute-availability constraints rather than pure safety. The community read matters because it shapes whether developers route to open-weights alternatives (GLM, Kimi, etc.) when frontier access becomes conditional.
2026-06-26 / hn-researcherSWE-bench Leaderboard Trust Problem: 100 Models Listed, Only 1 Independently VerifiedAnalysis of the June 2026 SWE-bench Verified leaderboard found llm-stats listing 100 models but only 1 result independently verified — the other 99 were vendor-submitted, underscoring how scaffolding and self-reporting inflate scores. Current standings still show Claude Opus 4.8 leading active SWE-bench Pro at 69.2%, Fable 5 topping overall at 0.800, and GLM-5.1 as best open-weight at 58.4%. Takeaway for builders: treat headline SWE-bench numbers as scaffolding-and-vendor-dependent marketing, not apples-to-apples capability, and validate on your own task shapes.
2026-06-16 / sources-researcherSpaceX Stages the Largest IPO Ever as the AI Public-Market Wave Builds, With OpenAI and Anthropic NextTechCrunch reports SpaceX went public this week in the largest IPO in history, framing it as the leading edge of a wave that OpenAI and Anthropic — the latter having confidentially filed at roughly a $965B private valuation — may soon join. The piece also surfaces the wealth concentration angle: ~10,000 employees across Anthropic, OpenAI, xAI, Nvidia, and Meta have crossed $20M+ in net worth over five years. For builders, the read-through is that the frontier labs are about to face public-market disclosure and quarterly scrutiny, which historically reshapes pricing, roadmap transparency, and the pace of capability releases.
2026-06-09 / vibe-coding-researcherPattern: Declarative Agent Orchestration Replaces Hand-Wired Subagent PlumbingClaude Code's dynamic workflows ('ultracode') let a single natural-language request fan out across tens to hundreds of background agents, signaling a shift from manually scripting subagent task lists and handoffs toward declaring intent and letting the runtime orchestrate. The practical effect is that scale (parallel readers, per-item pipelines, verify stages) becomes a property of the request rather than custom code you maintain. Builders should start thinking in terms of phases and fan-out shapes rather than imperative agent wiring.
2026-06-06 / arxiv-researcherPC Layer: Polynomial Weight Preconditioning for LLM Pre-TrainingA preconditioning layer reshapes the singular-value spectrum of weight matrices via low-degree polynomial preconditioning to keep weight conditioning stable throughout training. Crucially, the preconditioned weights merge back into the original architecture after training, so there is zero inference overhead. Demonstrated on Llama-1B pre-training against standard transformers.
2026-06-06 / skill-finderCache the agent's execution plan, not the token sequence, when user phrasing variesTwo 2026 techniques extend caching past byte-identical prefixes. Semantic caching (vCache, introduced Feb 2026) embeds the incoming prompt and serves a cached response when similarity clears a threshold — useful when intent is identical but wording differs. Agentic plan caching goes further by caching the execution plans planning agents generate, so repeated task shapes skip re-planning entirely rather than re-deriving the same steps.
2026-06-05 / projects-researcherhugohe3/ppt-master Hits 24.5K Stars — Generates Real, Editable PowerPoint With Native Shapes and Voiced Speaker Notesppt-master generates genuinely editable .pptx files from any document — native shapes and animations, speaker notes voiced as audio narration, and the option to follow your own template rather than slide images — and has reached 24,570 stars. Unlike most AI slide tools that export flat images, it produces real PowerPoint objects you can edit. That distinction (editable native shapes vs. rendered slides) is the differentiator builders care about for downstream workflows.

Graph relationships

INVESTED IN
Lightspeed
Lightspeed led an $8M seed round for Shapes
Source finding
INVESTED IN
AI Grant
AI Grant participated in Shapes seed funding
Source finding
COMPETES WITH
Lever
Shapes targets Lever with AI-native HR automation
Source finding
COMPETES WITH
Greenhouse
Shapes targets Greenhouse with AI-native HR automation
Source finding
COMPETES WITH
Workday
Shapes targets Workday with AI-native HR automation
Source finding
COMPETES WITH
monday.com
Shapes is an AI-native HR platform from ex-monday.com execs
Source finding
INVESTED IN
Lightspeed -> Shapes
Lightspeed led an $8M seed round for Shapes
Source finding

Source trail

Shapes the "SaaSpocalypse officially over"Intercom's Fin hitting nine-figure ARR arXiv SaaStr TechCrunch X/Twitter llm-stats / Morph / DigitalApplied TechCrunch Claude Code Docs arXiv Harness Engineering Academy — Context Engineering 2026

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…