How can the Insurance Industry best leverage Industry AI and unlock its augmented decision making potential in 2024 ?

As we wrap up 2023 I reflect on some of the learnings of 2023 in regards to the AI revolution which will transform every industry .

A recently published study by the UK Department of Education - The Impact of AI on UK jobs and training is a very insightful statistical based paper on which roles have the most potential for AI Augmentation.?

Report link here

Highlights of the report:

The Finance and Insurance sector is more exposed to AI than any other sector in the UK.

Of the top sixteen high automation occupations :brokers, pensions and insurance clerks and assistants ,call centre occupations, customer service occupations,other administrative occupations ,telephone salespersons ,financial administrative occupations , seven are common to Brokers, MGA's and Carriers .

The clear occupational group with the highest share of tasks exposed to AI technology are clerical jobs with about a quarter of tasks highly exposed to potential automation by AI. "Knowledge worker” exposure is only partial, suggesting a stronger augmentation potential and productivity benefits, rather than job displacement.

So the Insurance and Financial services industry has the most to gain from AI Productivity so why have they not fully embraced it yet ?

1. Insurance industry given its risk averse nature and heavy regulation is slow to embrace emerging technologies until it has a strong governance model in place .
2. Can Insurance employees Trust what the AI is telling them?
3. Does AI have the context of the insurance industry ?
4. Does the AI understand where I am in a company specific process for guidance ?
5. Generative AI is complex so how do they solve this technical Rubik's cube ?

When ChatGPT burst on the scene in October 2022 with the amazing ability to suggest very comprehensive answers and solutions to questions it was like the advent of the printing press except you the prompt writer could write the book with ChatGPT's help.

One year later a survey outlines where most enterprises are with AI in Production:?

The top 5 Motivations to pursue AI are : cost savings 39% , keeping up with trends 35%, genuine customer needs 33%, keeping up with competitors 31% and leadership interest 28%.

The top 5 pain points when deploying AI apps are : model output accuracy 39% , data security 33% , hallucinations 28%, associated costs 28%,model speed and performance 24%.

75-86% of large to small enterprises don't want to build , train and secure their own LLM models as they lack the in house AI skills and cyber skills to do so.

Our Insurance conversations have confirmed that some Insurers , Brokers and MGA's are experimenting with AI but seeing similar pain points with 40% bot accuracy and employees saying it's unable to help them with tasks as it has no context , they are seeing some success with Predictive AI and fraud use cases .?

The biggest issue facing the insurance industry we heard repeatedly is unstructured data in the form of complex PDF's during onboarding new clients and claims management that takes lots of underwriter and claims handler time up to 40% , OCR has been unable to handle anything beyond simple PDF.

Also the lack of data to reference on replacement of expensive Items or buildings and data augmentation and data cleansing by AI would be welcomed.

So what is Vertical Industry AI ?

Horizontal LLM are great at general knowledge questions as they can reference 280 billion plus parameters.

Industry AI from OpenEnterprise has two additional layers of training , the second layer is pre trained domain specific foundational industry models , industry knowledge graph and industry metadata.

The third layer is industry language models for retrieval augmented generation , we further train the model to the unique company level so that structured ,unstructured and 3rd party insurance data can be accessed by end users for very specific process coaching by industry Co-pilots and industry apps. These are Industry grounded in the specific processes to get the right data to the right user at the right time weather that be risk, accrual or underwriter data.?

Built by AI experts? providing purpose built Industry AI Models and apps to advance decision making ability by humans, scaling up the most valuable decision makers in enterprises, leading to exponential returns for early adopters .

Our vertical industry AI solutions have the power to datafy expert knowledge, meaning they can convert the knowledge stored in the minds of a company's experts into synthetic data. By tapping into the knowledge of insurance subject matter experts, an AI model can achieve more personalised and highly reliable outputs.

Click HERE to see a demo

OpenEnterprise.ai a salesforce ISV benefits for Insurance employees, partners and customers :

• Process specific agents for predictive, generative and conversational data apps and models?
• Secure and trusted Industry grounded retrieval augmented generation means no hallucinations or bias issues with 95% plus accuracy
• New hire onboarding reduced 50% as they have their own digital interns, helping with repetitive and time consuming tasks as they have the entire knowledge of the enterprise which has been datafied.
• Huge reduction in customer operations with 30-40% productivity gains for decision makers like underwriter and claims handler with the human in control?
• Provide pre trained foundational Vertical Industry AI models.
• Transform the CX and EX for customers , employees and partners by focussing on outcomes vs the current broken processes that hold enterprises back
• Industry AI can transform change management and compliance with more focus on outcomes and rapid prototyping and visibility of the key data used to make decisions for auditors and regulators.

OpenEnterprise.ai Benefits for customers :

• No rare AI? skills needed by customers; we provide admins with no code UX.??
• No infrastructure needed, we provide secure pre-trained VPC Sagemaker instances and train models for POC's , we will work with SI's on AI MSP, a new success model .
• Tiny Industry specific models are much more cost effective in token usage reducing running costs 10X.
• Leverage the Einstein Trust layer for Trusted AI, data intelligence and prompt engineering built into salesforce deployments.
• Data connectors directly into salesforce, Datacloud and other data warehouses.???
• All data is securely stored in salesforce which is the enterprise class leader for data security and trust is their number one value.?