Technology · 10.7

Enterprise Software

Quantified news impact based on deduplicated, quality-reviewed events in a single 20-day window.

Average net impact+19weighted per event
Average gross impact21absolute weighted average
Events16same period
Confidence72%model average
Raw intensity53before weighting
Previous period0net impact
Gross changeInsufficient base
Diffusion speed0.8 / day
Concentration9%
Direction disagreement12%
Source independence97%13 sources
IMPACT TREND

20-day effective impact

Positive 14 · Negative 1 · Mixed 1

Current periodPrevious period
EVENT EVIDENCE

Events ranked by effective impact

Page 2 of 2 · 16 events

Macro0

No evidence on this page.

Industries0

No evidence on this page.

Companies6
01
Companiespositive

Zhongwang Software Adjusts Share Buyback Plan: Fund Range Increased from 10-20 Million to 40-80 Million Yuan

Zhongwang Software announced an adjustment to its share buyback plan, increasing the total fund range from a minimum of RMB 10,000,000 (10 million) and a maximum of 20,000,000 (20 million) to a new range of RMB 40,000,000 (40 million) to 80,000,000 (80 million). The adjustment, which includes the 2000 figure in the upper limit, retains all other terms of the plan unchanged.

Effective
+13
Intensity
50
Confidence
85%
Horizon
Short term
02
Companiespositive

Anthropic Product Designer's Side Project Claude Design Goes Official, Surpasses 1 Million Users in First Week

Anthropic product designer Nate Parrott's side project Claude Design, a visual communication tool, has been officially adopted as a product. It surpassed 1 million users in its first week. The tool exports PDF, PPTX, HTML and integrates with Canva, Adobe, Gamma, and others. It is not an image generation model and requires users to import their own logos and brand assets.

Effective
+10
Intensity
40
Confidence
70%
Horizon
Medium term
03
Companiespositive

Tencent Cloud WorkBuddy Debuts on HarmonyOS PC Market as First Desktop Office Agent

Tencent Cloud announced that its WorkBuddy has been officially listed on the HarmonyOS PC app market, becoming the platform's first desktop office intelligent agent. This marks the initial availability of a dedicated desktop office AI tool on HarmonyOS, expanding Tencent's productivity ecosystem to Huawei's operating system.

Effective
+8
Intensity
40
Confidence
75%
Horizon
Short term
04
Companiespositive

D-Wave Quantum Expands AT&T Deal, Cuts Network Optimization from Hour to Under 15 Seconds

D-Wave Quantum and AT&T have expanded their cooperation, with D-Wave's quantum annealing technology reducing a network optimization workload from about one hour to under 15 seconds. AT&T plans to integrate the technology into its agentic AI tools, which helped reduce customer service downtime by approximately 12 million hours in 2025. The expanded deal will evaluate quantum annealing for fault detection, technician routing, network planning, and traffic management.

Effective
+6
Intensity
30
Confidence
50%
Horizon
Long term
05
Companiespositive

Haibo Sichuang and Others Establish Beijing Qingneng Lvdian Technology Co., Ltd.

Haibo Sichuang and other investors co-founded Beijing Qingneng Lvdian Technology Co. , Ltd. , a recently established company with a business scope encompassing artificial intelligence basic and application software development, software sales, software development, and machinery equipment sales. The company is jointly held by Haibo Sichuang and others.

Effective
0
Intensity
10
Confidence
30%
Horizon
Short term
06
Companiesneutral

Anthropic Reveals Claude Code Team's Internal Verification Loop with Four Automated Checks

Anthropic on July 22 disclosed the Claude Code team's internal verification loop, which runs four automated checks after code writing: code-review (bug fixing), simplify (redundancy reduction), verify (end-to-end), and design (visual alignment). This shifts the agent workflow from manual review to iterative auto-verification and repair. The team uses these Skills daily, with underlying support from built-in /verify, CLAUDE. md commands, and GitHub Actions. Anthropic also provided guidance on writing custom verification Skills, which can be triggered at four automation levels. The Skill format is becoming an open standard adopted by multiple vendors.

Effective
0
Intensity
30
Confidence
70%
Horizon
Medium term