B4 Index / The Continuum / July 22, 2026

The Continuum

B4 Research July 22, 2026

Edition 3

Week of July 15 to July 22, 2026 · 2 capability signals, 10 funding events, 1 papers

1,603

Categories tracked

0

Score movements

$4.0B

Capital tracked

2

Capabilities confirmed

1

Papers reviewed

Funding

10
CompanyRoundCategoriesAnnouncedAmount
AtomsSourceVentureJUL 22$1.7B
AtomsSourceEquity round (unlabeled)JUL 22$1.7B
MeshySourceSeries BJUL 21$400M
Candid HealthSourceSeries DRevenue Cycle Management (RCM)JUL 22$120M
ArrakisSourceSeries AAI Agent Frameworks & OrchestrationJUL 22$38M
NaturalSourceSeries APayments Processing+1JUL 20$30M
GrittSourceSeries AJUL 21$26M
InfinitySourceSeedMulti-LoRA Adapter Serving & Inference Optimization+1JUL 20$15M
CascadeSourceSeedConstruction Project Intelligence & Bid Lead DataJUL 22$3.5M
CascadeSourceSeedConstruction CRM & Pursuit Management+1JUL 21$3.5M

Capability signals

2
  • Production ProvenAgentic AI replacing legacy BI in production (Tradeshift / Amazon Quick)

    Evidence JUL 20 · Source

  • Production ProvenMulti-model AI architecture embedded in a data platform (Couchbase Capella iQ on Bedrock)

    Evidence JUL 20 · Source

Papers

1
Small Benchmark

Loading an agent skill does not keep its requirements active in long contexts — a code-audit task passed 8/10 runs in an ~11k-character clean context but only 3/10 at 299k characters (trend-level, p=0.0698), while a detailed external checklist passed 10/10 versus 5/10 for generic self-checking (p=0.0325); a second task passed all conditions, so no universal context-length threshold is supported.

When and How Context Rot Appears in Coding Agents: A White-Box Study of Agent Skills in Code Auditing

What it means For long agent sessions, externalize compliance into explicit checklists the agent must tick against rather than trusting a loaded skill or self-review to stay active; small run counts, so treat effect sizes as indicative.

arXiv preprint · JUL 20

Every row here is confirmed before it publishes, and the research is free. The full database and the score updates behind it are in a B4 subscription.