Composite insurer in Saudi Arabia goes full Agentic

Welcome back, HyperAgent. The world is moving fast; it is up to us to keep updated with the latest news and trends. But it's also up to you to keep yourself educated on how to work with the latest news and trends. This week, we’ll begin looking at how and where insurance professionals can place themselves in an increasing AI-augmented world.

Today’s Insights

• Finding your place in an AI-augmented world

• Face-scans detects early cancer risks

• It’s happening - insurers are going full Agentic

• Organize your physical and digital desktop with a picture and a prompt

AI FOR INSURANCE PROFESSIONALS THIS WEEK

Composite insurer in Saudi Arabia goes full Agentic

Amana Cooperative Insurance has rolled out an AI-first, end-to-end motor-claims journey - without changing a single line of core-system code. Using a swarm of master-and-worker AI agents plus RPA “hands,” the Riyadh-based insurer now pushes many claims straight through from First Notice of Loss (FNOL) to settlement and fraud checks, all while policy-holders chat on WhatsApp or e-mail instead of waiting on hold.

Why this matters

• Core freeze, growth squeeze. Amana’s board set an aggressive 2025 motor-portfolio target but banned any tweaks to legacy systems during a tech-modernisation freeze. The team needed digital speed without the usual multi-year core upgrade.

• Agentic AI meets RPA. Large-language-model agents handle dialogue, decisions and orchestration; RPA bots write data back into the core and partner platforms. The result keeps the solution “non-intrusive” yet fully automated.

How the flow works

1. Smart FNOL intake. Policy-holders upload documents via portal, e-mail or WhatsApp; an AI agent checks completeness, fetches accident data through APIs, and confirms receipt in seconds.

2. Validation & triage. Worker agents run technical and statistical rules, trigger challenges when estimates look high, and hand suspected fraud to specialists.

3. Real-time status. Throughout the journey, chat agents keep customers - and brokers - updated on claim status via their preferred channel.

4. Straight-through settlement. When no rule is violated, the claim auto-settles; RPA posts reserves and payments directly into the core.

Under the hood

Early wins

• Faster FNOL: Document completeness rates up, reducing re-work and call-backs.

• Lower leakage: Automated cost-analysis challenges inflated estimates before approval.

• Customer glow: Always-on chat beats IVR queues, boosting NPS.

Quantitative KPIs are under NDA, but early dashboards already show material cycle-time cuts and higher straight-through-processing percentages.

Signal to watch: Amana plans to push the same agent stack into sales, CRM and even workshop management next. Keep an eye on how quickly Saudi policy-holders come to expect “WhatsApp-to-settlement” as the new normal.

The solution was built and delivered by eData, a regional expert in process automation through and with Agentic AI.

AI agents are coming for your busywork—here’s how to lead, not lag, in the digital labor era

Salesforce CEO Marc Benioff says we’ve officially entered the “agentic age”—where AI doesn’t just assist but actively works as a digital employee. From customer support to executive planning, Salesforce is embedding AI agents into every layer of business, powered by tools like Einstein, AgentForce, and the new agent-driven Tableau.

🏢 Why should insurance professionals care?

This is a turning point from AI as a tool to AI as a team member. Insurers relying on large service centers, analytics teams, or operations staff will see core workflows shift. Underwriting, claims triage, fraud detection, and policy servicing are now prime candidates for automation by autonomous agents, not just scripted bots or static models.

💼 What this means for your current job

If you manage tasks, analyze data, or handle repetitive decisions, those tasks will soon be outsourced to AI agents. But this doesn’t eliminate your role—it elevates it. You’ll oversee how agents perform, define the logic they follow, and ensure they align with compliance and customer expectations.

Mastering digital labor workflows makes you future-proof. Being someone who orchestrates AI agents—instead of competing with them—puts you on track for leadership in automation governance, agent deployment, and innovation roles. Think of it as a new management layer: not over people, but over digital teams that scale your impact.

CUTTING-EDGE AI

Show me your face, and I…

Is face scanning the new health check? A Boston research team turned 58,851 everyday photos into a deep-learning tool called FaceAge. Feed it a selfie and it spits out your “biological age” — a truer read on health than birthdays. In hospital tests on 6 196 cancer patients, doctors who added FaceAge to their usual eyeball test jumped from 61 % to 80 % accuracy in spotting who would still be alive six months later. Patients whose faces scored “85 +” did worst, proving the model is a fast, low-cost flag for hidden risk. 

Why insurance professionals should care

• One-click pre-checks. Ask applicants for a selfie and instantly spot people who look unusually “old” for their age, then order extra exams only when it matters.

• Sharper pricing. Blend FaceAge scores into your underwriting model to segment healthy-looking 55-year-olds from frail-looking ones and trim manual review time.

• Early-warning outreach. Run the tool on your in-force book to find members aging faster than expected and nudge them toward screenings or lifestyle coaching. 

Try it on yourself here.

How to get ready

• Try a sandbox pilot: pair five underwriters with a FaceAge API and compare quote speed and decisions against a control group.

• Study the data basics: learn what drives biological-age models, from skin texture to facial symmetry, so you can explain results to colleagues.

• Tap your partners: invite insurtechs and reinsurers to a two-hour workshop on facial-age use cases and share quick-win ideas.

• Build a mini-roadmap: outline where FaceAge slots into your quote flow, what extra data you’ll need, and how success will be measured (time-to-quote, hit ratio, loss ratio).

Start small, learn fast, and you’ll be ready when customers expect a selfie to be part of every health-cover conversation.

THE INSURANCE AI ACADEMY

Confident prompts can make chatbots lie - here’s how to avoid it

When you tell a chatbot, “Keep it super short—I’m 100 % sure,” you’re nudging it to invent facts. A new study of popular models shows that tight word limits boost wrong answers by as much as 20 %, while an over-confident tone makes some bots 15 % less likely to correct you. Small models wobble most; larger ones like Claude 3.7 and Gemini 1.5 stay steadier. 

How to cut down hallucinations

• Ask for sources, not silence.

  Example: “Give a two-sentence answer and add one trusted link.”

• Keep questions neutral.

  Instead of: “I’m sure Paris is in Italy—right?” try “Where is Paris? Please verify.”

• Choose sturdier models for serious work.

  Example: Use Claude 3 or Gemini 1.5 for compliance reports.

• Set a “truth over brevity” rule.

  System note: “If unsure, say so or ask a follow-up rather than guessing.”

• Double-check key outputs.

  Example: Run a quick web search or have a colleague scan the answer before sending.

YOUR CAREER, YOUR FUTURE

Learn-it-all wins in an AI-first market - news from SXSW

AI now memorizes every clause and tariff, so the real edge for insurance professionals is the ability to learn quickly, connect ideas across disciplines, and judge the machine’s output with a critical eye. Breadth plus discernment is the new career insurance.

In the past, a motor underwriter proved worth by recalling every deductible nuance from memory. Today, a large language model drafts that policy in seconds. What the model cannot do—at least not yet—is understand the client’s evolving risk context, reconcile conflicting data, or explain trade-offs to a nervous broker. Those tasks demand broad perspective, deep thinking, and the confidence to challenge the algorithm.

The same shift echoes across the value chain. Claims teams already see AI flagging anomalies faster than any human audit. Actuaries watch models generate price scenarios on the fly. Brokers use chatbots to quote mid-market cyber cover at midnight. In every case, the technology removes rote work and exposes a new gap: professionals who can interpret, question, and refine the machine’s first draft.

Your personal growth playbook

Deep thinking beats blind trust

Strong outputs demand sharp human judgment. Treat the model’s draft as a smart first pass, then put your experience to work: check the logic, test unusual scenarios—think multi-jurisdiction claims, fast-evolving cyber risks, or sudden regulatory tweaks—and challenge anything that feels off.

Your grasp of policy nuance, local law, and real-world loss patterns turns raw AI text into sound, insurable language. Remember, the model is powerful because you understand what good looks like; your critical eye is what keeps clients protected and the business profitable.

AI will keep raising the waterline. Those who swim in many currents—technical, creative, and relational—and who think deeply about what the machine delivers will not just stay afloat; they will chart the course for insurance’s next wave.

👀 AI Agent of the week: Prompt optimizer

Our Agent of the week will guide you to the perfect prompt; click here and answer the questions, then copy / paste the generated prompt into your LLM of choice, and see the magic of a well-designed prompt happen

PRODUCTIVITY TOOLS AT HOME AND AT WORK

Money Motivator: The new iPROMISE app is turning procrastination into progress by making you bet on yourself. Users create challenges and put their cash on the line – if you skip that workout or miss that study session, you'll pay up. Perfect for anyone whose New Year's resolutions never make it to February.

Kompas: Conduct in-depth research across hundreds of webpages, saving you time by filtering out irrelevant data.

RFP on steroids: Trampoline AI helps you and your team craft winning RFP and keeps track on the success rate of them all, ensuring your next RFP has all chances of winning.

10 prompt templates for higher productivity: AI researcher Nelly R Q shared her top 10 prompts for turning ChatGPT into a thought clarifier, a reply generator, and more.

PROMPT OF THE WEEK

Organize your office, home , or computer desktop

How to Use:

1. Take a screenshot or photo of the area you want to organize physical or digital area)

2. Upload the image into ChatGPT

3. Copy and paste the prompt into the chat box

4. Send the prompt

Prompt: I've uploaded an image. First, identify what you're seeing. Then, suggest a better way to organize it, aiming for something that’s quick, efficient, and easy to maintain. Finally, provide a step-by-step plan or grouping strategy I can follow (if applicable).

WHAT’S TRENDING

LinkedIn announced a new AI-powered job search tool allowing users to find career opportunities that match their dream roles using natural language commands.

Brain surgery in 3D: A viral video is giving a peek into modern medicine. It shows doctors using augmented reality and tractography (a 3D modeling technique) to visualize the brain in real time during surgery, improving precision and safety.

AI Agents in the hospital: China has opened the world’s first AI-powered virtual hospital featuring 42 AI doctors across 21 departments. This facility can autonomously manage the entire patient care process, from diagnosis to follow-up, and has achieved a 93.06% accuracy rate on the MedQA dataset.

AI IMAGES OF THE WEEK

Where is the place for the humans in an AI world?

💡 That’s all folks - tell me what you think

Your feedback makes me create better content for you! 🙏🏼

Until next week, Frederik, eData & the AI Agents

Your growth,
your career,
your future