At a glance
Management Summary
Context of study and methodology
Exploring AI in the Swiss EAM sector
Swiss External Asset Managers are currently deploying AI along their value chain. However, empirical evidence about AI use cases, their potential and their constraints to improve processes in the Swiss EAM sector specifically is rare.
This study is exploring AI adoption in the EAM value chain. It sheds light on associated opportunities and restrictions for the EAM sector and for Fintech start-ups supporting EAMs.
- 41 Swiss EAMs participated in a structured online survey from April to June 2026.
- 9 Swiss EAMs and their advisors participated in semi-structured, qualitative video-interviews of 60 minutes each from March to July 2026.
Innosuisse study to analyze innovation opportunities
This study is funded by the Swiss Innovation Agency «Innosuisse» to inform sector participants and their suppliers about context and requirements of AI applications for the Swiss EAM sector.
It is conducted by the Lucerne University of Applied Sciences and Arts on behalf of – and in cooperation with – Finup, a Swiss WealthTech startup.
Main results
EAMs are exploring AI’s potential, however, AI integration into processes remains shallow
- All 41 EAMs engage with AI in some form, none reports having neither plans nor interest. Half already use or actively explore AI to enhance business processes, even if unsystematically and not deeply integrated.
- Expected efficiency and quality improvements are main drivers of AI adoption as is the desire to keep up with technological developments. Gaining more time with clients is the superordinate goal.
Adoption is two-speed: language-heavy tasks show traction, workflows lag behind
- In language- and content-heavy processes, such as investment research, proposal generation and client reporting, current AI use is already visible.
- By contrast, workflow- and data-heavy processes, such as client onboarding and data quality & reconciliation, show lower current AI use despite perceived AI potential or need for improvement.
- This suggests that adoption is constrained less by lack of interest and more by integration effort, data readiness and confidentiality requirements.
The gap is EAMs’ readiness, not AI’s potential
- To date, EAMs’ individual and organizational capabilities to fully leverage AI’s potential are a bottleneck. EAMs are deliberately cautious in moving forward:
- 83% consider AI a strategic opportunity;
- but only 27% consider their data sufficiently structured for AI;
- internal expertise and clear responsibilities also score low (41% each).
What EAMs need next is a strategy and a blueprint, not more tools
- EAMs ask for systematic education and a market overview of what exists and what it costs. Efficiency gains are hard to evidence while costs are immediate. The proof-point gap keeps budgets cautious.
- Barriers are manageable: data protection dominates (85%); regulatory concern fades with experience (52% among AI explorers & discussers vs. 30% among users), while reliability concerns persist across all groups.
(See sidebar, «Methodology: study design and sample».)
Survey results
AI adoption and interest in AI
AI adoption by EAMs
Most EAMs are already using or actively exploring AI, although adoption is often unsystematic.
Survey results
EAMs were asked to assess their firm’s current position regarding AI.
EAMs demonstrate an open mindset toward AI:
- Half of surveyed EAMs already use AI across selected processes (49%).
- A further 39% are currently exploring or planning AI initiatives.
- Only a small minority has discussed AI without concrete plans (12%).
- No respondents report having neither plans nor interest in using AI.
Further insights
However, current AI usage appears scattered and not deeply integrated into processes.
- According to interviewees, EAMs experiment with AI and explore its potential, but do not yet allocate significant resources to these efforts.
- AI engagement is typically driven bottom-up by selected individuals and is rarely institution-wide. One exception is that management often monitors selected AI risks. For example, some EAMs have formulated AI policies and guidelines to limit operational risks.
- Nevertheless, the development or implementation of a company-wide AI strategy remains rare.
- Interestingly, some EAMs engage in client interactions triggered by EAM clients’ own usage of AI, e.g., when clients ask specific questions regarding their portfolio holdings.
View data
| Position | Share |
|---|---|
| We already use AI across selected processes | 49% |
| We are currently exploring or planning AI initiatives | 39% |
| We have discussed AI, but no concrete initiatives are planned | 12% |
| We currently have neither plans nor interest in using AI | 0% |
Interest in AI
EAMs are interested in AI for efficiency and process quality, not primarily for firm positioning.
Survey results
EAMs were asked for their motivation and interest in using AI.
Streamlining tedious, labour-intense processes is top of mind:
- Potential efficiency gains are by far the strongest motivation for using AI, mentioned by 93% of respondents.
- Quality improvement of existing processes and keeping up with technological developments are also important drivers, mentioned by around three quarters of respondents (73% and 68%, respectively).
- Positioning as an innovator or market leader is clearly less relevant, mentioned by only 32% of respondents.
- Motivation differs by AI maturity: AI users more often mention quality improvement (90%) than AI explorers and discussers (57%), while efficiency gains are highly relevant for both groups.
Further insights
Increasing quality time with clients is the superordinate goal.
- According to interviewees, EAMs expect that AI may provide high potential to improve selected processes in the medium to long term.
- EAMs hope that AI will help them increase quality time with clients, but only few EAMs are currently able to provide details on concrete use cases that might help them reach this goal.
- Some interviewees caution against expecting AI to be a panacea for all types of operational and advisory challenges.
The figure compares motivations for using AI by firms’ current AI position. AI users are EAMs that stated they already use AI across selected processes. AI explorers and discussers include EAMs that stated they are currently exploring or planning AI initiatives or have discussed AI without concrete initiatives.
Reading example: Potential efficiency gains were mentioned by 93% of all respondents as a motivation for using AI. Among AI users, the share is 95%. Among AI explorers and discussers, it is 90%.
View data
| Motivation | Total | AI users | AI explorers & discussers |
|---|---|---|---|
| Potential efficiency gains | 93% | 95% | 90% |
| Quality improvement of existing processes | 73% | 90% | 57% |
| Keeping up with current technology developments and building capability | 68% | 70% | 67% |
| Positioning as innovator / market leader | 32% | 35% | 29% |
Survey results
AI use cases and AI readiness
Improvement areas vs. state of automation and AI potential
AI potential is rated high for many processes that EAMs wish to improve.
Survey results
EAMs were asked to identify processes that offer room for improvement. For each selected process, they assessed the current level of automation and AI potential*.
Improvement demand and AI potential point to a focused shortlist:
- Client onboarding, KYC / AML due diligence and investment research are the most frequently selected improvement areas, each named by at least half of surveyed EAMs.
- These processes are mostly described as fully manual or digitally supported, with client onboarding and KYC / AML due diligence particularly often still fully manual.
- AI potential is also assessed as high for these three processes.
- In general, AI potential is strongest where processes involve language, documents, research or client communication. However, low automation alone does not imply high AI potential: trade execution and financial & tax planning show comparatively lower AI potential.
Further insights
Whether high AI expectations can be met depends on process-specific factors: «manual» does not equal AI-ready.
- To date, some processes are not streamlined end-to-end, e.g., trade execution, where different actors (clients, EAMs, banks) as well as multiple interfaces, may hinder realization of AI potential.
- Some improvement areas are not process specific but still have AI potential, especially tedious, language-based tasks such as:
- writing newsletters or structuring complex email correspondence with clients;
- using AI for translations, transcripts or identifying specific mutual funds;
- using AI as a coach, e.g., when preparing client meetings or assessing funds;
- generating investment content, supporting stock selection or capturing back-office documents.
Processes are ranked by improvement demand, measured by the number of EAMs selecting the process as an improvement area (N). The left wing shows the reported automation level of the selected process; the right wing shows whether AI potential is assessed as high, limited or absent.
Reading example: Client onboarding & account opening was selected by 25 EAMs as requiring improvement. Of these, 7 describe the process as fully manual, 8 as digitally supported and 10 as partially automated; 1 assesses AI potential as absent, 6 as limited and 18 as high.
View data
| Process | N | Fully manual | Digitally supported | Partially automated | Fully automated | No AI potential | Limited | High |
|---|---|---|---|---|---|---|---|---|
| Client onboarding & account opening | 25 | 7 | 8 | 10 | 0 | 1 | 6 | 18 |
| KYC / AML due diligence | 21 | 6 | 7 | 7 | 1 | 0 | 3 | 18 |
| Investment research | 20 | 3 | 13 | 3 | 1 | 0 | 3 | 17 |
| Data quality & reconciliation | 18 | 2 | 7 | 7 | 2 | 0 | 4 | 14 |
| Proposal generation | 17 | 2 | 9 | 6 | 0 | 0 | 4 | 13 |
| Client reporting & communication | 16 | 0 | 7 | 9 | 0 | 0 | 2 | 14 |
| Regulatory compliance & documentation | 16 | 2 | 7 | 6 | 1 | 0 | 4 | 12 |
| Portfolio monitoring & risk management | 15 | 3 | 11 | 1 | 0 | 0 | 2 | 13 |
| Fee calculation & billing | 14 | 3 | 2 | 9 | 0 | 1 | 4 | 9 |
| Internal accounting & financial reporting | 14 | 3 | 7 | 4 | 0 | 0 | 5 | 9 |
| Ongoing client due diligence | 13 | 4 | 2 | 6 | 1 | 0 | 3 | 10 |
| CRM & client data management | 12 | 3 | 0 | 9 | 0 | 0 | 4 | 8 |
| Rebalancing & mandate management | 11 | 2 | 3 | 6 | 0 | 0 | 2 | 9 |
| Trade execution & order management | 11 | 5 | 3 | 3 | 0 | 2 | 5 | 4 |
| Investment & suitability compliance | 10 | 0 | 3 | 7 | 0 | 0 | 3 | 7 |
| Investment strategy & asset allocation | 8 | 0 | 8 | 0 | 0 | 0 | 3 | 5 |
| Portfolio implementation | 6 | 2 | 3 | 1 | 0 | 0 | 3 | 3 |
| Financial & tax planning | 3 | 1 | 2 | 0 | 0 | 0 | 2 | 1 |
* Based on three survey questions:
1) Please select those processes, which currently require the most effort and resources and that you wish to improve.
2) How would you rate the level of digitalization and automation of the selected processes?
3) How would you assess the potential of AI in these processes?
Current AI use and maturity level
AI adoption is concentrated in a few use cases and mostly remains experimental.
Survey results
EAMs that already use AI in selected processes (N=20) were asked to indicate their specific areas of application and the current maturity level of AI use.
- AI use among EAMs is concentrated in a limited number of processes and mostly remains experimental. Across all AI applications (65 in total), 66% are experiments with public tools, 23% are pilot use cases and only 11% are scaled implementations. Scaled use is concentrated in four processes that are closely linked to content generation.
- Investment research clearly stands out: 14 out of 20 EAMs already use AI in this area, mainly through experimentation with public tools or pilot use cases. Equity screening and qualitative/quantitative analyzes support fundamental research and active investment decisions.
- AI is also used relatively often in KYC/AML due diligence and client reporting and communication, each mentioned by eight EAMs. Notably, client reporting & communication shows the strongest signs of scaled or production-level implementation.
Further insights
Broader AI integration into core operational processes remains limited.
- General-purpose AI tools are mainly used to experiment and learn, e.g., chatbots (ChatGPT), AI assistants (MS Copilot, Claude), real-time research engines (Perplexity).
- Finance-specialized AI tools also play a role, e.g., Bloomberg’s AI chatbot (BloombergGPT) to analyze corporate data, portfoliochat.AI to analyze portfolios or low-budget proprietary pilot projects, e.g., AI layering on Excel.
- EAMs use mixed approaches for AI licensing: Some use corporate licences, some use individual licences, free offerings or subsidised trial offerings. Flat-fee chatbot subscriptions are common; token-based billing, e.g., for agentic AI use, is not mentioned.
- EAMs are still assessing whether use cases should be supported by highly specific AI tools, an integrated «super app», which agentic AI might be able to provide, or by embedded AI in standard software applications of established providers.
The figure ranks processes by how often they were named by EAMs already using AI. Each horizontal bar represents one process; the total bar length shows the number of EAMs using AI in that process, while the colored segments show the maturity level of AI use: experimentation with public tools, pilot use cases or scaled/production-level implementation.
Reading example: In investment research, 14 EAMs already use AI. Of these, 8 are experimenting with public tools, 4 use AI in pilot use cases and 2 report scaled or production-level implementation.
View data
| Process | Experimentation | Pilot | Scaled | Total |
|---|---|---|---|---|
| Investment research | 8 | 4 | 2 | 14 |
| KYC / AML due diligence | 7 | 1 | 0 | 8 |
| Client reporting & communication | 5 | 0 | 3 | 8 |
| Proposal generation | 5 | 1 | 1 | 7 |
| Portfolio monitoring & risk management | 2 | 4 | 0 | 6 |
| Reg. compliance & documentation | 2 | 2 | 0 | 4 |
| Investment strategy & asset allocation | 2 | 0 | 1 | 3 |
| Client onboarding & account opening | 2 | 0 | 0 | 2 |
| Data quality & reconciliation | 1 | 1 | 0 | 2 |
| Internal accounting & financial reporting | 2 | 0 | 0 | 2 |
| Ongoing client due diligence | 2 | 0 | 0 | 2 |
| Rebalancing & mandate management | 1 | 1 | 0 | 2 |
| Financial & tax planning | 2 | 0 | 0 | 2 |
| Fee calculation & billing | 1 | 0 | 0 | 1 |
| Investment & suitability compliance | 1 | 0 | 0 | 1 |
| Portfolio implementation | 0 | 1 | 0 | 1 |
Bringing it all together: improvement priorities and AI potential vs. current adoption
Several high-priority processes show strong AI potential, but current AI use is still concentrated in only a few areas.
Survey results
Improvement demand and AI potential point to similar processes:
- KYC/AML due diligence, investment research, proposal generation, client reporting and data quality & reconciliation combine high improvement demand with high perceived AI potential. Client onboarding stands out: it is the most frequently cited pain point, but perceived AI potential is slightly lower than for other priority processes.
- Rebalancing, portfolio monitoring and ongoing client due diligence show high perceived AI potential but are less frequently prioritized for improvement.
- Trade execution, financial & tax planning and portfolio implementation show comparatively low improvement demand and lower perceived AI potential.
Current AI use is concentrated:
- It is high for investment research, KYC/AML due diligence and client reporting & communication, while some processes show a potential–adoption gap, especially data quality & reconciliation and proposal generation: both show strong AI potential, but current AI use remains limited
Further insights
What the map means for decision-makers:
- For EAMs, good starting points are processes where improvement demand, AI potential and current adoption already overlap, e.g., investment research, KYC/AML due diligence and client reporting. However, interviews show that budgets may stall because efficiency gains are invisible, while costs are visible. Early AI use cases should therefore deliver measurable wins.
- For vendors, opportunities lie in processes where AI potential is high but current use is still limited, such as proposal generation and data quality & reconciliation.
- For operational processes such as onboarding, KYC/AML, regulatory documentation and data reconciliation, AI may need to be combined with workflow redesign, better system interfaces and data readiness rather than implemented as a stand-alone tool.
Hover a bubble to identify the process. Full values in “View data”.
- Client onboarding & account opening
- KYC / AML due diligence
- Investment research
- Data quality & reconciliation
- Proposal generation
- Client reporting & communication
- Regulatory compliance & documentation
- Portfolio monitoring & risk management
- Fee calculation & billing
- Internal accounting & financial reporting
- Ongoing client due diligence
- CRM & client data management
- Rebalancing & mandate management
- Trade execution & order management
- Investment & suitability compliance
- Investment strategy & asset allocation
- Portfolio implementation
- Financial & tax planning
Each bubble represents one process in the EAM value chain. Bubble size indicates the number of EAMs that already use AI in the respective process.
The horizontal axis shows the share of all surveyed EAMs selecting the process as requiring improvement.
The vertical axis shows the share of those EAMs assessing AI potential as high. AI potential was assessed only for processes previously selected as requiring improvement.
View data
| Process | Named by (of 41) | High AI potential | Firms using AI |
|---|---|---|---|
| Client onboarding & account opening | 25 | 72% | 2 |
| KYC / AML due diligence | 21 | 86% | 8 |
| Investment research | 20 | 85% | 14 |
| Data quality & reconciliation | 18 | 78% | 2 |
| Proposal generation | 17 | 76% | 7 |
| Client reporting & communication | 16 | 88% | 8 |
| Regulatory compliance & documentation | 16 | 75% | 4 |
| Portfolio monitoring & risk management | 15 | 87% | 6 |
| Fee calculation & billing | 14 | 64% | 1 |
| Internal accounting & financial reporting | 14 | 64% | 2 |
| Ongoing client due diligence | 13 | 77% | 2 |
| CRM & client data management | 12 | 67% | 0 |
| Rebalancing & mandate management | 11 | 82% | 2 |
| Trade execution & order management | 11 | 36% | 0 |
| Investment & suitability compliance | 10 | 70% | 1 |
| Investment strategy & asset allocation | 8 | 63% | 3 |
| Portfolio implementation | 6 | 50% | 1 |
| Financial & tax planning | 3 | 33% | 2 |
* We provide examples of potential use cases in the Appendix II.
Barriers to AI adoption
Data protection and confidentiality concerns are the dominant barrier to AI adoption.
Survey results
EAMs were asked to indicate barriers for AI adoption.
- Data protection and confidentiality concerns are by far the most frequently mentioned barrier, cited by 85% of EAMs.
- Concerns about reliability and accuracy of AI results as well as regulatory uncertainty are also relevant barriers, mentioned by 56% and 41% of respondents, respectively.
- EAMs generally see enough relevant AI use cases: only 2% cite limited relevant use cases as a barrier to AI adoption.
- Differences by AI maturity are visible mainly in regulatory uncertainty: AI explorers and discussers mention regulatory uncertainty more often (52%) than AI users (30%). By contrast, concerns about reliability and accuracy remain relevant for both groups (60% among AI users and 52% among AI explorers and discussers). Data protection and confidentiality concerns are dominant across both groups.
Further insights
EAMs monitor selected AI risks and costs but believe these are manageable.
- EAMs are careful to protect client confidentiality and ring-fence internal client data.*
- They scrutinize how AI connectors interact with current infrastructure and how this could affect business continuity. Auditability of AI and achievability of AI results are also important requirements.
- Potential supplier dependencies/vendor lock-in and fast technological change are perceived as challenges.
- Regulatory uncertainty and dealing with FINMA supervision are not seen as major AI barriers, as EAMs see themselves as experienced and prudent risk-takers.
- Cost-benefit considerations, however, are an issue: Interviewees appear more sensitive to the potential total costs of AI than the online survey results suggest. Examples include future running costs (incl. increased IT and insurance costs) and costs of institutional transformation.
- Overall, barriers to AI adoption are seen as manageable challenges rather than major roadblocks.
* Public AI models with servers abroad are seen as a specific risk.
The figure compares barriers to AI adoption by firms’ current AI position. “AI users” are EAMs that stated they already use AI across selected processes. “AI explorers and discussers” include EAMs that stated they are currently exploring or planning AI initiatives or have discussed AI without concrete initiatives.
Reading example: Data protection and confidentiality concerns were mentioned by 85% of all respondents, 80% of AI users and 90% of AI explorers and discussers.
View data
| Barrier | Total | AI users | AI explorers & discussers |
|---|---|---|---|
| Data protection / confidentiality concerns | 85% | 80% | 90% |
| Concerns about reliability and accuracy of AI results | 56% | 60% | 52% |
| Regulatory uncertainty | 41% | 30% | 52% |
| Cost-Benefit relation is not obvious | 17% | 15% | 19% |
| Limited relevant use cases | 2% | 5% | 0% |
AI readiness
EAMs see AI as strategically relevant, but internal readiness lags behind.
Survey results
To assess AI readiness, we asked EAMs about their AI capabilities and governance.
A large majority considers AI a strategic opportunity: 83% agree or strongly agree. However, execution readiness drops when it comes to specific details:
- Basic implementation conditions are assessed relatively positively:
- 66% agree or strongly agree that their IT infrastructure allows AI integration.
- 56% are willing to allocate budget to AI pilot projects.
- Organizational readiness is more mixed:
- 41% report clear responsibilities and processes for AI implementation.
- 41% report sufficient internal expertise to work with AI.
- Data readiness is the weakest area: only 27% agree that their data is sufficiently structured and of high quality for AI use. This gap may turn out being a major constraint for exploiting AI potential based on internal data.
Further insights
Individual and organizational readiness are a bottleneck to firm-wide AI deployment.
- EAMs are somewhat ambivalent on AI strategy and appropriate implementation:
- Important requirements for AI implementation are not yet fully in place, such as internal expertise, responsibilities and data governance.
- EAMs have high AI expectations but are not yet addressing implementation challenges strategically; instead, they are deliberately cautious in moving forward.
- EAMs want to gain AI expertise, but do not yet make it a high priority:
- Some EAMs seek external AI advice and collaborate with existing or new providers to improve step-by-step and keep up with current AI developments.
- Other EAMs plan to build AI capabilities by educating staff on AI systematically or by recruiting new talent. Indeed, AI education was mentioned in most interviews as an important topic, including both AI possibilities and AI-related risks.
View data
| Statement | 5 = strongly agree | 4 | 3 | 2 | 1 = strongly disagree |
|---|---|---|---|---|---|
| We consider AI a strategic opportunity for our firm | 44% | 39% | 12% | 5% | 0% |
| Our IT infrastructure allows for integration of AI solutions | 29% | 37% | 27% | 5% | 2% |
| Our data is sufficiently structured and of high quality for AI use | 15% | 12% | 54% | 15% | 5% |
| We are willing to allocate budget to AI pilot projects | 29% | 27% | 29% | 10% | 5% |
| We have sufficient internal expertise to work with AI | 17% | 24% | 29% | 27% | 2% |
| We have clear responsibilities and processes to support AI implementation | 17% | 24% | 24% | 29% | 5% |
AI adoption should start with a clear strategy, not tools.
Appendix
Appendices
Appendix I — List of interviewees
Coralie Bachmann-Gostanian, Relationship Manager, swisspartners, Zurich, 08.06.2026.
Kristian Bader, Board member, Fontaris AG, Bern, 20.03.2026.
Sara Bassi, Wealth Manager, Colombo, Zurich, 28.05.2026.
Nicole Curti, Managing Partner/CEO, Capital Y, Geneva, and President of Alliance of Swiss Wealth Managers (ASWM), 08.06.2026
Dimitri Petruschenko, PETRUSCHENKO CONSULTING, Zurich, 31.03.2026
Stefan Rammelmeyer, CEO, Winvest Asset Management, Zurich, 29.05.2026.
Oliver Riedel, Head of Business Operations & Digitalization, SoundCapital, Zurich, 10.06.2026.
Moritz Solèr, Founder and CEO, Emerald Wealth Partners, Zurich, 13.05.2026.
Dr. Jonas Steinmann, Portfolio Management & Operations, Managing Director, Invethos, Bern, 10.06.2026.
Appendix II — Priority AI use cases across the EAM value chain
- 1. Client onboarding & account opening1
- 2. KYC / AML due diligence2
- 3. Ongoing client due diligence
- 4. CRM & client data management
- 5. Financial & tax planning
- 6. Investment research3
- 7. Proposal generation4
- 8. Client reporting & communication5
- 9. Investment strategy & asset allocation
- 10. Portfolio implementation
- 11. Rebalancing & mandate management
- 12. Trade execution & order management
- 13. Investment & suitability compliance
- 14. Portfolio monitoring & risk management
- 15. Regulatory compliance & documentation (e.g., flow of funds)
- 16. Data quality & reconciliation (e.g., with custodian banks)
- 17. Fee calculation & billing
- 18. Internal accounting & financial reporting
1 Client onboarding & account opening
| Use case | Description | Ease of implementation | Data sensitivity |
|---|---|---|---|
| Document extraction & pre-fill | Extract and validate client data from identity, tax and legal documents, pre-fill account-opening forms and cut re-keying errors. | Quick win | Medium |
| Agreement generation & review | Draft custodial and service agreements from templates, auto-flag non-standard clauses and route internal approvals. | Core build | Medium |
| Automated client profiling | Build investor profiles and risk questionnaires from financial documents, stated goals and risk tolerance. | Core build | High |
| Agentic onboarding orchestration | Orchestrate the full chain (data collection, conflict resolution, e-signatures, custodian registration) autonomously. | Frontier | High |
2 KYC / AML due diligence
| Use case | Description | Ease of implementation | Data sensitivity |
|---|---|---|---|
| KYC extraction & SoW/SoF narratives | Pull data from identity and corporate documents, flag gaps, draft source-of-wealth / source-of-funds narratives for review. | Core build | High |
| Flow-of-funds documentation | Auto-generate transaction commentary for AML files, extract key facts from supporting documents. | Core build | High |
| Adverse media screening | Continuously screen news and public records, produce decision-ready documentation packs per client. | Quick win | Medium |
| Sanctions & watchlist triage | Score alerts against sanctions and PEP lists, surface genuine hits and cut manual review load. | Quick win | Medium |
3 Investment research
| Use case | Description | Ease of implementation | Data sensitivity |
|---|---|---|---|
| Research synthesis & summarization | Digest market research, analyst reports and macro trends from internal and third-party sources into briefing-ready summaries. | Quick win | Low |
| Fund & product document analysis | Extract fees, risk disclosures and terms from prospectuses, factsheets and term sheets for side-by-side comparison. | Quick win | Low |
| Investment memo generation | Draft advisory notes and memos combining portfolio analytics with narrative market context, for human review. | Core build | Medium |
| Research co-pilot on firm knowledge | Natural-language interface over the firm's own research base, with source-linked, archivable answers. | Core build | Medium |
4 Proposal generation
| Use case | Description | Ease of implementation | Data sensitivity |
|---|---|---|---|
| Investment proposal drafting | Generate personalized proposals from risk profile, objectives and preferences, including suitability rationale. | Core build | High |
| Pitchbook auto-generation | Populate on-brand decks with portfolio data, charts and commentary drawn from internal sources. | Quick win | Medium |
| Product suitability comparison | Compare products on fees, risk and performance, generate justification notes ready for the proposal. | Quick win | Low |
| Meeting pack assembly | Compile client goals, portfolio data, market updates and disclosures into a structured, client-ready pack. | Quick win | Medium |
5 Client reporting & communication
| Use case | Description | Ease of implementation | Data sensitivity |
|---|---|---|---|
| Automated portfolio commentary | Generate personalized performance narratives from risk analytics, market outlook and client preferences. | Core build | Medium |
| Multi-language communication | Draft client e-mails and reports in the client's language, adapting tone, style and terminology. | Quick win | Medium |
| Call & meeting summarization | Transcribe client meetings, extract decisions and action items, and file follow-ups to the CRM. | Quick win | Medium |
| Natural-language portfolio Q&A | Let advisors, and eventually clients, query portfolios in plain language with explainable answers. | Frontier | High |
Quick win: up and running in weeks with ready-made tools and minimal IT effort – value visible in the same quarter.
Core build: a managed implementation over several months, connected to the firm’s own systems and data – where the durable gains sit.
Frontier: emerging, largely autonomous AI, promising but unproven – for most firms one to watch, not yet to build.
About
About the authors
An Innosuisse innovation project, conducted by the Lucerne University of Applied Sciences and Arts in cooperation with Finup.