Assessment of Investment Sustainability in AI Application Sector and Analysis of High-Potential Niche Areas for Long-Term Growth

Based on the latest market data and case studies, I will systematically analyze the investment sustainability assessment framework for the AI application sector and the niche areas with long-term potential for you.

• Revenue Growth Quality
  : Examine whether revenue comes from real customer payments rather than one-time projects or government subsidies. Take Geneplus, an AI medical company, as an example: its 2022 revenue reached RMB1.815 billion (driven by COVID-19 testing dividends), but plummeted to RMB473 million in 2023, a year-on-year decline of 73.9%, exposing the risk of relying on short-term opportunities [1]. Truly sustainable applications should show stable growth in core businesses, such as rebounding to RMB557 million in 2024 (17.8% year-on-year), indicating that core businesses have regained a foothold [1]
• Gross Margin Trend
  : Reflects productization level and pricing power. Geneplus’ gross margin increased from 41.9% in 2022 to 68.5% in H1 2025, indicating that its product structure is adjusting toward high value-added directions [1]
• Customer Retention and Expansion
  : Includes Net Dollar Retention (NDR), customer churn rate, etc.

• Cash Flow Status
  : Geneplus’ operating cash flow in H1 2025 was RMB-34.78 million, with only RMB96.01 million in cash on hand, but net current liabilities reached RMB1.775 billion, facing significant capital pressure [1]
• Clarity of Profit Path
  : The Moore Threads case shows that even with revenue of RMB785 million in the first three quarters, it still had a net loss of RMB724 million, with an expected annual loss of RMB730 million to 1.168 billion, and the profit timeline is unclear [2]
• R&D Investment Efficiency
  : Matching degree between R&D expense ratio and revenue growth

• Core Algorithms and Data Advantages
  : Whether it has exclusive data or algorithms
• In-depth Scene Understanding
  : Accumulation of know-how in vertical industries

• Market Share and Growth
  : For example, in China’s short video market, Douyin and Kuaishou occupy nearly 70% of user time share, forming a duopoly pattern [3]
• Depth of Moat
  : Network effects, switching costs, etc.

• AI enterprises are more suitable for growth stock valuation models such as forward P/E ratio and PEG [2]
• HSBC Investment Management pointed out that current AI valuations are “seemingly high, but far from the extreme level of the dot-com bubble burst” [4]
• Need to be alert to irrational speculation: Moore Threads’ cumulative increase reached 723.49% in 5 trading days after listing, with valuation soaring from RMB300 billion at the beginning of the year to RMB4.5 trillion, expanding nearly 20 times in a year, far exceeding performance support [2]

• PS/PCF vs Growth Rate
  : For unprofitable enterprises, it is necessary to evaluate together with revenue growth quality
• Implied Growth Rate
  : Whether the future growth expectations implied by valuation are achievable

• The RaaS model proposed by BaiRong CloudGroup emphasizes “delivering results rather than functions”, representing the evolution of AI business paradigm [5]
• Three-layer value exchange model: pricing by task, post salary system, revenue sharing based on value creation [5]
• Core advantages: Elevate AI from “tool” to “productivity” level, realizing “measurable usage, measurable effect, and closely linked revenue and value” [5]

• Market Pain Points: Traditional recruitment faces three dilemmas: “slow hiring, inaccurate assessment, and difficult diagnosis”
• AI Solutions: BaiRong CloudGroup’s “BaiCai” product reduces the recruitment cycle from 28 days to 2 days, increases resume success rate from 60% to 90%, and expands HR’s single-position recruitment volume from 5 to 20 [5]
• Long-Term Logic: Rigid demand for cost reduction and efficiency improvement due to continuous rise in labor costs

• Application Scenarios: Professional service links such as finance, taxation, legal affairs, HR, etc.
• Typical Case: BaiRong CloudGroup’s “BaiJian” platform compresses the feedback cycle of cross-border professional services from 90 days to 14 days, reduces project costs from RMB3-6 million to within RMB1 million, using the “9:1 Silicon-Carbon Collaboration” (90% of work done by AI) [5]
• Market Space: The digital penetration rate of China’s enterprise service market is still low, with huge room for improvement

• Market Size: China’s short video users reach 1.04 billion, with a usage rate of 93.8%, ranking first in the online audio-visual application segment for 6 consecutive years [3]
• AI Empowerment Value: Kuaishou fully empowers online marketing services through AI, realizing comprehensive efficiency improvement from content understanding, recommendation distribution to delivery decision optimization [3]
• Commercial Achievements: Kuaishou’s “Keling AI” has accumulated revenue exceeding RMB700 million since its launch in June 2024 [3]
• Long-Term Logic: After the fade of traffic dividend, refined operation becomes a rigid demand

• Typical Case: Kuaishou E-commerce’s GMV reached RMB1.08 trillion, growing 16% year-on-year, and AI technology significantly improved the efficiency of matching people and goods [3]
• Competition Pattern: Meituan still faces loss pressure in fierce competition, highlighting the value of AI for cost reduction and efficiency improvement [6]

• Market Opportunities: Demand driven by population aging and unbalanced medical resources
• Challenges and Risks: From the Geneplus case, AI medical enterprises need to prove their commercialization capabilities, and the market has higher requirements for profit paths [1]
• Investment Points: Choose niche areas with clear payment意愿 (B-end hospitals/C-end patients), rigid scene demand, and clear supervision

• Typical Enterprise: Kutai Smart focuses on personalized intelligent customization of clothing, building a C2M business ecosystem [7]
• Core Value: Realize large-scale personalized customization through AI, subverting the traditional “inventory-driven” model
• Long-Term Logic: Dual drive of consumption upgrade and manufacturing upgrade

• Application Scenarios: Intelligent risk control, quantitative investment advisory, intelligent claims settlement, etc.
• Advantages: High data standardization, strong payment意愿, measurable effect in the financial industry
• BaiRong CloudGroup Case: Has served more than 8,000 institutional customers, covering multiple industries such as finance, internet, retail, etc. [5]

• Representative Product: Kuaishou’s “Keling AI” maintains a leading position globally, and has been widely used in professional creative scenarios such as advertising marketing, e-commerce, film and television, short dramas, animation, etc. [3]
• Commercial Progress: Launched the 2.5 Turbo model in September 2025, which has greatly improved in text response, dynamic effects, style retention and other dimensions [3]
• Market Space: The content industry scale is trillions, and AIGC can reconstruct the content production and distribution model

• Product Case: BaiRong CloudGroup’s “BaiZhi” product reduces the delivery cycle of in-depth reports from 20 days to 4 days, improving delivery efficiency by 80% [5]
• Demand Pain Points: Fragmented information acquisition, lack of methodology in knowledge production, long delivery cycle

• Market Dynamics: OneRobotics (Shenzhen home robot manufacturer) listed in Hong Kong, raising USD210 million [8]
• Technical Trend: Meta acquired Singapore-based AI agent company Manus, accelerating the integration of AI in the intelligent hardware field [9]
• Long-Term Logic: Demand from rising labor costs and aging society

• Industry Stage: Evolving from L2 to L3/L4, with accelerated commercialization landing
• Investment Points: Focus on technical routes (pure vision vs lidar), commercial progress (Robotaxi/mass-produced vehicles), and industry chain position

• Technical Iteration Lag
  : For example, Moore Threads prompts that “new-generation AI chip products have not yet generated revenue, and multiple links are required from certification to mass production, with uncertainties such as technical iteration lag and customer onboarding failure” [2]
• Wrong Technical Route Selection
  : May face the risk of being subverted

• Customer Acceptance Lower Than Expected
  : B-end customers have long decision-making cycles, and trial-to-paid conversion rate is low
• Increased Competition
  : For example, Meituan’s takeaway business faces fierce competition from Alibaba, losing 15-20 percentage points of market share [6]
• Macroeconomic Fluctuations
  : Affect enterprise IT expenditure budget

• Continuous Losses
  : Most AI application enterprises are in a loss state, relying on external financing [1][2]
• Cash Flow Tension
  : For example, Geneplus faces the pressure of net current liabilities up to RMB1.775 billion [1]
• Valuation Correction Risk
  : Current valuation partially overdraws future growth expectations

• Judgment Criteria: Whether it solves real pain points, whether there is existing payment意愿, whether it can generate measurable ROI
• Avoid purely conceptual projects without business closed loop

• Sustainability of revenue growth (non-one-time projects)
• Gross margin improvement trend (reflecting productization capability)
• Customer retention and expansion data

• Diversify by track: Enterprise services, medical care, finance, manufacturing, etc.
• Diversify by stage: Early exploration + growth stage + mature stage
• Diversify by business model: Subscription model, pay-per-use model, revenue sharing based on results, etc.

• Quarterly tracking: Revenue, gross margin, cash flow, customer retention
• Annual assessment: Technical iteration progress, market share changes, competition pattern evolution

• Priority selection of enterprises with clear profit paths and rapid revenue growth
• Alert to concept stocks with pure speculation and no fundamental support

• Use market sentiment fluctuations to deploy high-quality targets in batches when they correct
• Pay attention to but be cautious about targets after IPO surge (such as Moore Threads) [2]

• AI application enterprises with technical platform capabilities and cross-industry replication
• For example, BaiRong CloudGroup’s “Result Cloud” platform supports the deployment of silicon-based employees in multiple business scenarios [5]

• In vertical fields such as AI+medical care, AI+finance, AI+manufacturing, select enterprises with deep industry know-how
• Pay attention to scaling opportunities brought by the increase in industry penetration rate

• RaaS (Result as a Service) model represents the evolution direction of AI business paradigm [5]
• Choose enterprises that can deliver measurable business results and are deeply tied to customer value

1. Revenue growth rate >30% (core business non-one-time projects)
2. Gross margin >60% and showing an upward trend
3. Net Dollar Retention (NDR) >120%
4. Positive operating cash flow or rapid improvement
5. R&D expense ratio 20-40% (need to match revenue growth)

1. Have core algorithms or data barriers
2. Products have cross-industry replication capabilities
3. Increase in depth and frequency of customer use

1. Founding team has dual background in industry and technology
2. Clear equity structure, no over-concentration
3. Standard corporate governance, transparent information disclosure

AI application sector investment has long-term sustainability, but short-term speculation risks need to be alert. Investment sustainability assessment should focus on:

1. Commercial Maturity
   : Revenue growth quality, gross margin trend, customer retention and expansion
2. Financial Health
   : Cash flow status, clarity of profit path, R&D investment efficiency
3. Technical Barriers and Differentiation
   : Core algorithms/data advantages, in-depth scene understanding
4. Valuation Rationality
   : Combine with growth indicators (PEG), avoid irrational speculation

:

• AI+Enterprise Services (EX): Intelligent recruitment, professional service automation
• AI+Customer Experience (CX): Intelligent marketing, e-commerce and live streaming
• AI+Vertical Industries: Finance, medical care, manufacturing (C2M model)
• AI+Content Creation (AIGC): Video generation, enterprise knowledge production
• AI+Intelligent Hardware and Robots: Home service robots, autonomous driving

Investment Recommendations:

. At the same time, maintain portfolio diversification, dynamically track core indicators, and deploy high-quality targets when valuation is reasonable, which is expected to obtain excess returns in the AI application wave.

---

[1] Yahoo Finance - “Post-Pandemic Dividend Fades: Geneplus Stays Steady and Moves Forward Again” (https://hk.finance.yahoo.com/news/疫情紅利退潮-吉因加穩扎穨打再出發-003317403.html)

[2] Yahoo Finance - “723% Surge in Five Trading Days! Moore Threads Urgently Issues Trading Risk Warning” (https://hk.finance.yahoo.com/news/五個交易日暴漲723-摩爾線程緊急提示交易風險-誰在幕後推高-063006974.html)

[3] Yahoo Finance - “【Smart Stock Selection】Strengthen AI to Boost Momentum: Kuaishou Worth Accumulating at Low Levels” (https://hk.finance.yahoo.com/news/智識選股-強化ai增動能-快手趁低收集-175500678.html)

[4] Yahoo Finance - “HSBC Investment Management Says AI Valuations Not at Extreme Bubble Levels Like the Dot-Com Era” (https://hk.finance.yahoo.com/news/匯豐投管指-ai-估值未現科網極端泡沫-220949086.html)

[5] Yahoo Finance - “Deliver Results with AI: BaiRong CloudGroup Launches RaaS Strategy and ‘Result Cloud’ Platform” (https://hk.finance.yahoo.com/news/用ai-交付結果-百融雲創發布raas戰略及-結果雲-平台-105700043.html)

[6] Yahoo Finance - “Post-Earnings: Comprehensive Summary of Major Banks’ Latest Investment Ratings on Meituan After Quarterly Results Announcement” (https://hk.finance.yahoo.com/news/績後-文綜合大行於美團-03690-hk-公布季績後最新投資評級-044945549.html)

[7] Baidu Encyclopedia - “Qingdao Kutai Smart Co., Ltd.” (https://baike.baidu.com/item/青岛酷特智能股份有限公司/19985096)

[8] TechMeme - “Shares of Shenzhen-based OneRobotics opened flat in their HK debut” (http://www.techmeme.com/251229/p25#a251229p25)

[9] CNBC - “Meta acquires intelligent agent firm Manus, capping year of aggressive AI moves” (https://www.cnbc.com/2025/12/30/meta-acquires-singapore-ai-agent-firm-manus-china-butterfly-effect-monicai.html)