Week 12 facilitating by

Atharva & Saqib

2025-10-17 02:53:32 UTC

It will require Data Standardization. (Data Fragmentation is a major issue )
Clear Data protocols are required and they must align with company's workflow.
They Should be able to tackle Data security and privacy
AI hallucinations - May leak data to other organizations

Yutong Wu 24858098

24858098 — 2025-10-17 02:53:56 UTC

I think the JPMC approach can be generalized by other organizations if they have enough resources and preparations for the implementation. JPMC invested lots of investment to the internal AI, and built reliable data security control and robust governance. These preparations reduced the overall risk of implementing Gen AI of JPMC.

Hui

2025-10-17 02:54:26 UTC

accuracy: when the suggestion is presented to the customer, the answer should be constructive and useful to face customer feedback.
Responsibility: the consultation gives the suggestion, however, the decision is made by the customer. The financial loss's responsibility percentage.
Financial monitor: The suggestion it gives will be monitored by the financial institutions.

2025-10-17 02:55:00 UTC

1. For smaller companies lacking experienced AI teams, building such systems may prove prohibitively expensive.

2. Integrating multiple modules can create dependencies between them, leading to difficulties in updating, slower response times, and increased resource usage.

Freya

2025-10-17 02:56:30 UTC

technology maturity: Ensure the security of the cloud and API

data security: ensure customer data is completely isolated, strictly authenticated, and operates within the bank's internal framework.

VAISHNAV SAILESHKUMAR (24410746)

vaishnavsk2001 — 2025-10-17 02:57:18 UTC

Before it upgrad coach from consulatant assistant to direct to customer they just need to meet some standards in data security, accuracy,responsibility , and transparency. Customers data must be protected.

Hann

2025-10-17 02:57:58 UTC

Not easily.

JPMC has a large scale deployment and unique data access. It invested 17 trillion in technology and employs a lot of technologists and engineers. Other organizations are hardly replicate.

Akshai

2025-10-17 02:58:08 UTC

Yes, It is possible if there’s strong data governance, internal expertise, and regulatory compliance but without clear accountability frameworks and cybersecurity measures it is risky.

Atharva & Saqib

2025-10-17 02:58:11 UTC

Accuracy & Reliability of Advice
Regulatory & Compliance Control
Transparency in every recommendations
Security & Privacy (Data Isolation)

Ira

2025-10-17 02:58:30 UTC

Data privacy and security: To prevent client information leaks.

Had to reach near-zero error rates through trials and traceable data sources.

Chances of risk need for continuous monitoring, with human oversight for high-risk transactions.

Transparency is necessary to maintain customer trust.

The system must enter, monitor and save all AI-client communications for expert review.

Stephanie Wang

love999stephanie — 2025-10-17 02:58:36 UTC

Data protection and confidentiality: No client inputs or records will be leaked, and all processing must remain within JPMC’s secure systems.

Trustworthy and verified knowledge: Coach must only provide accurate, validated information from approved JPMC sources.

Akash Durairaj & Miranda

24367785 — 2025-10-17 03:00:05 UTC

1. Follow all banking laws and compliance rules.

Start with voice-assisted or hybrid models before full rollout.
Client data privacy should never be at risk.
Keep human checks and monitoring.

Akshai

2025-10-17 03:00:19 UTC

Upgrad must maintain uncompromising standards in content quality , user experience , and data privacy when moving to a direct to customer model. It’s vital that automation still provides reliable guidance and keeps learner trust.

Miranda, Akash

2025-10-17 03:00:56 UTC

Not all organizations can safely ‘copy’ JPMorgan’s comprehensive approach. Their system works well because of its size, money, and strict control. Other companies usually don’t have that level of infrastructure or capital at their disposal.

Without the same rigorous oversight applied by JP (strict data-security rules, clear risk management and testing, trained staff) , other companies can face data leaks, wrong outputs, and compliance issues.

We think that JPMC’s model show what’s possible, but most companies would need simpler and smaller steps before building such complex multi-model AI systems. Still, firms can learn from JPMC’s ideas, learning from them by starting small with pilot project / ‘crawl-walk-run’, use human review for AI outputs, and let staff experiment with guardrails.

2025-10-17 03:01:29 UTC

Service Quality and Consistency. No matter who the customer is or when and where they engage, service quality and user experience must remain the same and top-quality.

Yihan Chen

2025-10-17 03:02:45 UTC

For the organisations with complex structure has multiple businesses and massive data, I think they can generalised the JMPC approach, but they should notice that the JPMC approch is based on the situation of JPMC, it is not suitable for all the organisations, so the best way is they use it for reference.

And for those small-size organisations with simple business, it is not necessary to generalise it.

Mayur(24550706)

2025-10-17 03:04:50 UTC

The other organizations cannot completely generalize it as it depends on their scale and resources. but the organizations can generalize their architecture principles.

The main reasons why it can't be replicated:

high budget $17 billion: compliance and data frameworks