For buyers watching AI reshape software.
Inaugural issue: the SaaS counter-attack landed, outcome-based pricing crossed verticals, and the system-of-record moat showed its age.
Week 21 of 2026 · May 23, 2026
Big read
Welcome to The Application Layer. The publication exists because the layer above the model is shifting faster than the model layer itself. Where The AI Stack Weekly tracks the cross-stack flywheel (software / hardware / networking / capital) and The Model Pulse drills the model layer (lineage / benchmarks / vendor signals), The Application Layer covers what enterprise buyers actually deploy — vertical agentic packages from frontier labs, SaaS-incumbent counter-attacks, vertical-AI startups, and the pricing-model shift that follows. The editorial thesis: SaaS-as-application gives way to data-as-application; the model becomes the runtime; the data cloud becomes the platform; the application is increasingly ephemeral.
W21 was unusually loud on this dimension. Workday Q1 FY27 (May 21) printed the first concrete agentic-ARR number from a back-office SaaS incumbent — approaching $500M with 4,000-plus customers on Workday-built agents (more than doubled QoQ). CEO Aneel Bhusri pledged to hold FY27 headcount flat by deploying Workday's own agents internally. SaaS investors should re-baseline Salesforce and ServiceNow agentic-revenue models off that print; Bhusri's flat-headcount commitment is the canonical SaaS-CEO statement that AI substitution is now an operating-model commitment, not a thought experiment.
The system-of-action thesis moved down the stack in two places at once. Workday Sana announced agents for ITSM and travel-and-expense — pulling ticket-resolution, onboarding, and T&E workflows into the HR system-of-record using Workday's existing identity and policy graph (early adopter 2H 2026, GA late 2026). Salesforce Agentforce Coworker (May 21) shipped inside Slack, Teams, Salesforce, and ChatGPT search bars — bundled free into existing Agentforce Enterprise / Unlimited / Agentforce 1 tiers, structurally pressuring per-seat AI-assistant pricing across the universal-search category. Both are system-of-action moves that displace the 'vertical SaaS for X' tier; CIOs running parallel Copilot, Glean, or Now Assist pilots should pause incremental seat expansions until they have benchmarked the Coworker / Sana shape.
Pricing structure shifted on three independent vectors inside one window. Zendesk Relate 2026 (May 19, Denver) made outcome-based pricing literal — $1.50 per dual-verified automated resolution ($2.00 PAYG), with the resolution verified first by the resolving AI agent and then by an independent evaluation model. CEO Tom Eggemeier framed it as 'the era of the chatbot is over' and explicitly priced agents as 'a unit of labor.' On the lab side, Anthropic's Agent SDK June 15 cutover (announced May 14) moves Claude Code, GitHub Actions, and third-party agents off subscription rate limits onto separate $20-$200/mo metered credit at API list prices — a 12x-175x effective price increase per workload, ending Claude Code subscription arbitrage. Google's Antigravity 2.0 + Gemini 3.5 Flash bundle (May 19) lifted Flash list pricing to $1.50 / $9 per Mtok (3x predecessor) and packaged the agent-runtime as a premium SKU. Architects renewing SaaS contracts this quarter should add an outcome-based clause negotiation to their playbook; finance leaders modeling FY27 token spend should treat agent-runtime tokens as premium, not commodity.
The vertical-specialist cohort emerged as the durable insurgent against horizontal generalists. Lexroom closed $50M Series B (May 19, Left Lane Capital) eight months after Series A on a civil-law Europe thesis distinct from Harvey's common-law US/UK lineage — legal AI is now jurisdiction-specific, not category-wide. Vi Healthcare closed a $145M secondary-plus-primary at a $1.64B valuation (May 19) and shipped a vertical AI agent suite GA with disclosed $2B-plus measured value across 100-plus enterprise customers — the first vertical-AI unicorn in healthcare with audit-grade outcome metrics. Dust closed $40M Series B (May 18, Sequoia and Abstract co-led, strategic checks from Snowflake and Datadog) on a multi-agent-collaboration thesis: knowledge work consolidates around multiple agents talking to each other, not isolated chatbots. Anthropic's 290,000-employee Hitachi deal (May 18) plus the Frontier AI Deployment Center is the largest disclosed enterprise commitment to a frontier lab to date. Sonar acquired AI-native code-review startup Gitar (May 21) — engineering tooling RFPs should now consolidate code review plus verification into a single procurement decision.
Watch list for the next 7-14 days: Salesforce Q1 FY27 (May 28) is the next inflection — first agentic-revenue color from the largest CRM incumbent post-Workday's $500M anchor. Microsoft Build (June 2-3) will reveal whether MAI ships as a vertical-specific line or stays horizontal. Adobe Q2 FY26 (mid-June) tests Firefly Agents adoption across creative workflows. Anthropic's 60-day Glasswing update (early-to-mid June) tests whether capability-gated procurement becomes a durable product category. For CIOs renewing in 2H 2026, the operating reality has changed: SaaS list-price increases now need to clear an outcome-based-pricing alternative or get repriced.
Vertical movements
2026-05-22 · Anthropic · Security · Frontier Lab · Usage Based
Project Glasswing — Claude Mythos Preview for critical-infrastructure defense
CISOs who deferred a Glasswing decision at the April launch should re-open the file: the May 22 update confirms partners have identified large numbers of high- and critical-severity zero-days across systemically important open-source projects, and Anthropic is opening enrollment to allied-government partners and additional critical-software maintainers. If your sector is on the critical-software map and you are not in the coalition, your defenders will face an asymmetric vulnerability-disclosure curve as Mythos-class capability propagates to other labs.
2026-05-19 · Anthropic · Operations · Frontier Lab · Usage Based
Claude Managed Agents — self-hosted sandboxes + MCP tunnels
CIOs whose architecture review boards had blocked Claude Managed Agents over data-perimeter and shadow-AI risk now have a direct answer: tool execution can run inside customer infrastructure via Cloudflare / Daytona / Modal / Vercel, and private MCP servers reach the agent via outbound-only encrypted tunnels with no public-internet exposure. The 18-month CISO veto on production agent deployments lost its load-bearing argument this week; revisit stalled Managed Agents pilots and pre-existing on-prem-only RFPs before the next steering committee.
2026-05-21 · OpenAI · Engineering · Frontier Lab · Seat Based
OpenAI Codex 26.519 — Goal Mode GA, Appshots, Plugin Sharing
Engineering leads waiting for an explicit GA signal on long-horizon autonomous coding now have it: OpenAI is officially standing behind hours-to-days agent runs against a single objective, and ChatGPT Business workspaces gain admin analytics on accepted-code rates and per-user adoption. Move at least one real refactor, migration, or test-coverage sprint onto Goal Mode this cycle; the gap against teams still pair-prompting at chat-turn granularity will compound through the rest of the quarter.
2026-05-20 · Cohere · Other · Open Weights · Usage Based
Cohere Command A+ — 218B MoE under full Apache 2.0
Regulated-industry CIOs, sovereign-cloud architects, and public-sector procurement officers who declined hosted-API frontier models now have a 218B-parameter MoE that runs on 2x H100 or 1x B200, generates native source-grounded citations, and carries no licensing tether or non-compete. Re-open any 'frontier capability is incompatible with our data-sovereignty posture' decision from the past 12 months; the open-weights frontier crossed the regulated-industries bar this week.
Sources cohere.com/blog/command-a-plus
2026-05-19 · Google DeepMind · Engineering · Frontier Lab · Hybrid
Antigravity 2.0 + Managed Agents API (Gemini 3.5 Flash runtime)
Engineering heads on Gemini CLI have a 28-day forced migration to the new agy CLI before June 18, and every dev-tools buyer has a Google-backed agent runtime with parallel sub-agents, scheduled background tasks, and a Managed Agents API that spins up isolated Linux sandboxes per call. Pair that with Gemini 3.5 Flash list-price at $1.50 / $9 per Mtok (3x predecessor) and Anthropic's Agent SDK billing split: the lab consensus this week is that agent-runtime tokens are now premium SKUs, not commodity inference — rebuild your FY26 token-spend forecast before the next budget review.
Sources blog.google/innovation-and-ai/technology/developers-tools/google-io-2026-developer-highlights
2026-05-22 · Microsoft · Operations · Frontier Lab · Usage Based
Microsoft Research Fara1.5 family — open-weight browser computer-use agents
Operations and shared-service heads automating booking, form-fill, procurement-portal, and account-driven workflows now have an open-weight browser-CUA stack that runs inside MagenticLite's Docker-isolated sandbox, pauses to ask on ambiguous or irreversible steps, and beats proprietary alternatives (27B at 72% vs Operator 58% and Gemini 2.5 CU 57%). If your operator-automation roadmap assumed only closed-weight CUA models were viable, schedule a parity bake-off this quarter — the deployment-control story (on-prem, no public internet) closes the last remaining enterprise objection.
Sources microsoft.com/en-us/research/articles/fara1-5-computer-use-agent
Incumbent responses
2026-05-21 · Workday
Workday Agents + Sana
SaaS investors got the first concrete agentic-revenue datapoint from a back-office incumbent — agentic ARR approaching $500M with customers on Workday-built agents more than doubling QoQ to 4,000-plus — and should re-baseline Salesforce and ServiceNow agentic-ARR models off this print. CIOs should treat Aneel Bhusri's pledge to hold FY27 headcount flat by deploying Workday's own agents as the canonical SaaS-CEO statement that AI substitution is now an operating-model commitment, not a thought experiment.
Sources Workday IR / SEC 8-K, CNBC
2026-05-21 · Workday
Sana for ITSM + Sana Travel Agent
Vertical-function heads renewing ServiceNow ITSM or SAP Concur in 2H 2026 should add Workday to the RFP — Sana for ITSM (early adopter 2H 2026, GA late 2026) and the Travel Agent (early adopter now) explicitly pull ticket-resolution, onboarding, and T&E workflows into the HR / finance system of record using Workday's existing identity and policy graph. Consolidation across employee-experience SKUs, not feature parity, is the agentic endgame this week's announcement makes operational.
Sources Workday Newsroom
2026-05-19 · Salesforce
Agentforce Life Sciences
Life-sciences CIOs evaluating vertical-CRM modernizations should pull forward Agentforce Life Sciences pilots — five-week go-lives at named pharma majors signal Salesforce is taking real share from the vertical-CRM incumbent in the HCP-engagement AI race ahead of the May 28 Q1 FY27 print. Investors should treat the 140-customer figure as the leading indicator of the Agentforce 'industry cloud' monetization thesis Benioff will defend on the call.
Sources Salesforce News
2026-05-21 · Salesforce
Agentforce Coworker (beta)
CIOs running parallel Microsoft Copilot, Glean, or ServiceNow Otto pilots should pause incremental seat expansions until they have benchmarked Coworker — it ships free to every existing Agentforce Enterprise / Unlimited / Agentforce 1 customer and unifies cross-system search without a separate contract. Per-seat AI-assistant pricing across the universal-search category is now structurally pressured, and Salesforce gains a 'system-of-action' wedge inside Microsoft and Slack surfaces it does not own.
Sources Salesforce News, secondary press
2026-05-20 · Snowflake
Cortex AI Function Studio
Data teams pricing Databricks Agent Bricks or building bespoke Mosaic AI pipelines for unstructured and multimodal workloads should defer those evaluations 60-90 days — AI Function Studio's automated benchmarking of prompts, models, and workflow strategies for document extraction, classification, and reasoning is Snowflake's first credible answer to Databricks' fine-tuning advantage, and it draws from existing Cortex credits with no new SKU.
Startup signals
2026-05-19 · Legal · $50M
Lexroom
Investors should treat legal AI as a jurisdiction-by-jurisdiction market, not a single category: Left Lane Capital led $50M (eight months after Series A) on a thesis that Harvey- and Legora-style common-law tooling cannot serve Europe's ~1M civil-law lawyers. GCs in Italy, Spain, Germany evaluating legal AI should add jurisdiction-native vendors to shortlists this quarter; founders building horizontal legal copilots face a structural data-infrastructure moat.
Sources Tech.eu
2026-05-19 · Healthcare · $145M
Vi
Health-system and pharma CIOs should benchmark vertically specialized AI agent platforms against horizontal LLM deployments before signing 2026 renewals: Vi reached unicorn status while disclosing $2B in measurable value across 100-plus enterprise customers, raising the procurement bar from pilot-stage promises to demonstrated ROI on patient navigation, physician next-best-action, and clinical-trial acceleration. Investors should note the General Atlantic / Pritzker / Square Peg syndicate as a template for late-stage vertical-AI capital.
Sources PR Newswire (Vi)
2026-05-18 · Other · $40M
Dust
CIOs piloting solo copilots should reconsider their architecture this week: Sequoia and Abstract co-led $40M Series B with strategic checks from Snowflake and Datadog on a thesis that knowledge work consolidates around multi-agent collaboration, not isolated chatbots. Platform decisions made in Q2 2026 will lock in agent topology for the next 24 months; founders building single-purpose copilots should plan for an interoperability layer or get acquired into one.
Sources SiliconANGLE
2026-05-18 · Operations
Hitachi + Anthropic
Industrial CIOs and frontier-model procurement leads should reset their vendor-concentration math: Anthropic now has the largest disclosed enterprise commitment to date (290,000 Hitachi seats, a 100,000-person AI-talent training pipeline, three-continent joint deployment center), stacked on PwC, KPMG, SAP, and BMS commitments inside the prior 30 days. Manufacturers benchmarking Lumada-style agent rollouts have a public reference architecture; OpenAI and Google competitors face a hardening Anthropic moat in regulated industrial accounts.
Sources Hitachi (primary press release)
2026-05-21 · Engineering
Sonar (acquiring Gitar)
Engineering leaders should consolidate AI coding tooling RFPs to include code review plus verification in a single procurement rather than running parallel pilots: Sonar (used by 75% of the Fortune 100 for code quality) absorbed Gitar's autonomous PR-fixing agent to extend SonarQube across the full agent-coding lifecycle. Standalone AI code-review startups without verification, governance, or CI-fixing capabilities face acquihire pressure within 6-12 months; AI-app founders should expect coding-agent margins to compress as bundles emerge.
Sources Sonar (primary press release)
Pricing shifts
2026-05-19 · Seat Based → Outcome Based
Zendesk
At Relate 2026 in Denver, CEO Tom Eggemeier announced an expanded outcome-based pricing model for the new Autonomous Service Workforce: customers pay $1.50 per committed automated resolution ($2.00 PAYG) only when the outcome is dual-verified — first by the resolving AI agent, then by an independent AI evaluation model — with spam and routine exchanges excluded. Eggemeier framed the shift as 'the era of the chatbot is over' and explicitly priced agents as 'a unit of labor.' For CIOs, this is the clearest signal yet that seat-based service software is being repriced as accountable AI capacity; competitors (Salesforce, ServiceNow) without a verification-and-evaluation framework face a procurement disadvantage when renewals come up.
Sources Diginomica (Relate 2026)
2026-05-14 · Seat Based → Usage Based
Anthropic
Effective June 15, Anthropic moves Claude Code, GitHub Actions, claude -p, and third-party agents (OpenClaw, T3 Code, Conductor, Zed) off Max-20x subscription rate limits onto separate $20-$200/mo metered credit at API list prices. Community math shows 12x-175x effective price increase per workload depending on token consumption, ending Claude Code subscription arbitrage. Architects running Claude Code at scale must model true API-rate burn before June 15 cutover or shift workloads — the cutover is the clearest pricing signal yet that agent-runtime tokens are premium SKUs.
Vertical scorecard
As of 2026-05-23
| Vertical | Leader | Challenger | Read |
|---|---|---|---|
| Legal | Harvey (common-law US/UK) | Lexroom (civil-law Europe, $50M Series B May 19); Claude for Legal partner ecosystem | Legal AI is now jurisdiction-specific, not category-wide; horizontal copilots face a structural data-infrastructure moat. |
| Healthcare | Vi ($1.64B unicorn May 19, $2B+ measured ROI across 100+ enterprises) | Abridge / Suki / Hippocratic AI; Salesforce Agentforce Life Sciences (140+ enterprises) | First vertical-AI healthcare unicorn with audit-grade outcome metrics; procurement bar moved from pilot-stage to demonstrated ROI. |
| Sales | Salesforce Agentforce (industry cloud + Coworker beta) | 11x.ai (digital workers); Clay (data enrichment); Apollo AI | Coworker bundle inside Slack / Teams / ChatGPT structurally pressures per-seat AI-assistant pricing across universal search. |
| Support | Zendesk Autonomous Service Workforce ($1.50 per verified resolution) | Sierra; Decagon; Cresta; Ada — and Salesforce Service Cloud Agentforce | Outcome-based pricing went literal in W21; competitors without verification-and-evaluation frameworks face procurement disadvantage on renewals. |
| Hr | Workday (agentic ARR ~$500M Q1 FY27; Sana ITSM + Travel) — system-of-record + system-of-action consolidation | Ema (universal employee); Maven AGI; BambooHR AI | Workday's flat-FY27-headcount pledge is the canonical SaaS-CEO commitment that AI substitution is operating-model, not optional. |
| Engineering | OpenAI Codex (Goal Mode GA + ChatGPT Business analytics) | Anthropic Claude Code (with Agent SDK June 15 repricing); Google Antigravity 2.0; xAI Grok Build; Cursor; Sonar+Gitar bundle | Agent-CLI race now four-way; harness portability is a procurement-clause issue, not a UX issue. |
| Security | Anthropic Claude Security + Project Glasswing (Opus 4.7-powered; 2,100 vulnerabilities patched in 3 weeks) | OpenAI GPT-5.5-Cyber (restricted access); CrowdStrike Falcon AI; Microsoft Security Copilot | Capability-gated procurement becomes a vendor-risk dimension; CISOs should pick a defender stack ahead of Mythos-class capability propagating. |
| Operations | Microsoft Fara1.5 + Anthropic Claude Managed Agents (sandboxes + MCP tunnels) | Dust ($40M Series B May 18, Sequoia + Abstract); OpenAI Codex computer-use; xAI Grok Build | Browser-agent + multi-agent orchestration moved into open weights this week; on-prem deployment story closes the CISO objection. |
Architecture watch
Outcome-based pricing crosses verticals in a single window
Three independent vectors (CX SaaS, frontier-lab agent runtime, frontier-lab inference) shifted pricing structure inside one window — and all three name the unit of work as the priced object (resolution, agent-run, premium agent-runtime SKU) rather than the seat. CFOs and procurement leads should add an outcome-based clause negotiation to every SaaS renewal this quarter; finance teams modeling FY27 token spend should treat agent-runtime tokens as premium, not commodity inference. Expect ServiceNow, Salesforce Service Cloud, and Workday to ship verification-and-evaluation frameworks within 90 days or cede pricing-narrative leadership.
- Examples
- Zendesk $1.50/verified resolution (Autonomous Service Workforce), Anthropic Agent SDK June 15 cutover (12x-175x effective), Google Antigravity 2.0 + Gemini 3.5 Flash 3x list-price reset
Sources Diginomica (Zendesk Relate 2026), Anthropic support, blog.google
System-of-action moves down the stack into adjacent system-of-record territory
Two SaaS incumbents (Workday, Salesforce) used the same week to push system-of-action features down the stack into adjacent system-of-record territory — Workday pulling ticket resolution and T&E into HR, Salesforce embedding agents in productivity surfaces it does not own. For CIOs running parallel Copilot / Glean / Now Assist pilots, the procurement question shifts from 'best AI assistant per surface' to 'which system-of-record consolidates the agent surface.' Per-seat AI-assistant pricing is structurally pressured; multi-system-of-record portfolios become harder to defend.
- Examples
- Workday Sana for ITSM + Travel (attacking ServiceNow / SAP Concur), Salesforce Agentforce Coworker inside Slack / Teams / ChatGPT (attacking Microsoft Copilot / Glean), Anthropic Hitachi 290K-employee deal (industrial Lumada displacement)
MCP + self-hosted sandboxes become the enterprise data-perimeter standard
Four announcements this week — three from frontier labs, one from a data-cloud incumbent — converged on the same architectural pattern: agent execution stays inside the customer's data perimeter, the data cloud (or sandbox runtime) becomes the platform, and the model is the runtime that runs against it. Security and platform teams should evaluate this as the new baseline for agent-deployment threat models; the 'data egress to vendor' objection that gated production agent deployments through Q1 2026 has a clear answer now.
- Examples
- Anthropic Managed Agents self-hosted sandboxes (Cloudflare / Daytona / Modal / Vercel), Anthropic MCP tunnels (research preview), Cohere Command A+ Apache 2.0 (2x H100 on-prem deployment), Snowflake Cortex AI Function Studio (data-cloud-native agentic primitives)
Vertical specialists displace horizontal generalists in regulated industries
Three closed transactions this week share a structural pattern: vertical specialists with deep domain data and outcome verification beat horizontal generalists in their verticals — including beating frontier labs' generic vertical bundles. Investors should treat the late-2025 'horizontal frontier models will eat verticals' thesis as falsified for regulated industries (legal, healthcare, industrial). Founders building category-agnostic copilots in regulated verticals face structural data-infrastructure moats and outcome-measurement gaps they cannot close with model capability alone.
- Examples
- Lexroom (civil-law Europe legal AI, $50M Series B), Vi (healthcare unicorn at $1.64B with $2B+ measured ROI), Anthropic-Hitachi 290K-employee industrial deal
Sources Tech.eu, PR Newswire, Hitachi PR
Watchlist
May 28
Salesforce Q1 FY27 earnings — first Agentforce revenue color post-Workday's $500M anchor
Workday's Q1 FY27 print (May 21) set the first agentic-ARR baseline ($500M approaching). Salesforce as the largest CRM incumbent will be asked on the call to provide a comparable disclosure. Investors should track: Agentforce ARR, customer count, deal sizes, and any flat-headcount commitment matching Workday's.
June 2 - 3
Microsoft Build 2026 (San Francisco) — MAI line + Foundry consolidation
Microsoft has been notably quiet through I/O week on its own MAI model line. Build is where MAI vertical packages, Copilot agent stack moves, and Foundry consolidation are likely to land. CIOs running Azure / Foundry pilots should hold model-selection decisions until after Build; vertical-AI founders should expect Microsoft to claim adjacent territory.
June 5 - 15
Anthropic Project Glasswing 60-day update + Mythos status
First Glasswing update (May 22) showed ~50 critical-infrastructure partners scanning. A 60-day update either expands enrollment to allied governments + critical-software maintainers or holds the current footprint. Either outcome conditions whether capability-gated procurement becomes a durable product category — and whether Mythos-class capability gets a permissive vs gated release path.
Mid-June
Adobe Q2 FY26 earnings — Firefly Agents adoption across creative workflows
Adobe shipped Firefly Agents in Q1 but has not disclosed ARR or customer counts. The Q2 print tests whether creative-workflow agentic adoption follows the customer-support / HR / sales pattern (rapid ARR ramp) or moves slower in creative use cases where the unit of work is harder to verify. Marketing and creative-ops procurement teams should anchor on this print before Q3 renewals.
May 26 - June 6
Vertical-AI startup funding wave continuation
W21 saw $235M close across three named vertical-AI rounds (Vi $145M, Lexroom $50M, Dust $40M). The follow-on question is whether the cohort raises in W22-W23 confirm a structural shift (vertical specialists > horizontal copilots in regulated verticals) or whether this was a one-window cluster. Watch Crunchbase / PitchBook for legal / healthcare / CX-specialist closes.
May 26 - June 13
Outcome-based pricing competitive response — Salesforce, ServiceNow, Adobe
Zendesk's $1.50 / verified resolution model (May 19) is the clearest outcome-based-pricing signal yet from an enterprise software vendor. Competitors without verification-and-evaluation frameworks face procurement disadvantage on renewals. Watch Salesforce Service Cloud, ServiceNow Now Assist, Adobe Firefly Agents for matching announcements within 30 days.
June 9 - 13
Apple WWDC 2026 — vertical Apple Intelligence + partner-model integrations
Apple absent from the vertical-packaging cycle through Q2. WWDC is the only major Apple disclosure window in 1H. Vertical-app developers and on-device-AI architects should expect Apple Intelligence repositioning and possibly partner-closed-model integration changes (Anthropic, OpenAI, Google) that re-condition vertical-app monetization on iOS.
Changelog
- Inaugural issue. The Application Layer launches as the third weekly publication, sibling to The AI Stack Weekly and The Model Pulse. Editorial frame: the layer above the model — vertical packages, incumbent SaaS responses, vertical-AI startups, pricing-model shifts. Voice rule: every block answers 'what should a CIO, vertical-function head, SaaS investor, or vertical-AI founder do differently this week because of this?'
- First-issue scorecard establishes 8 verticals as the baseline tracking surface (legal / healthcare / sales / support / hr / engineering / security / operations). New verticals will be added as the publication grows; rows will refresh weekly as leadership shifts.
- First-issue framing connects to the Pulse and Weekly: vertical packages from labs (Anthropic Glasswing, OpenAI Codex Goal Mode GA, Cohere Command A+, Microsoft Fara1.5) overlap with Pulse coverage but get reframed here through the procurement lens; incumbent SaaS earnings (Workday Q1 FY27) and vertical-AI startup raises (Vi, Lexroom, Dust) extend the Weekly's capital-flow coverage into the application layer.