Our AI stories on Oracle Fusion

How AI agents are working on Oracle Fusion Sales and Service, and the systems around them: the challenge, what the AI does, and what it changes. Customers are described by type of business.

Sales and opportunities

Agents that read tenders and emails, create and update opportunities, assess contract risk and learn from past bids.

Opportunity creation agent

The challenge

  • Business development teams read long tender documents and customer emails by hand.
  • The details are re-keyed into the CRM, which slows the first response and lets some enquiries slip.
  • Late entry delays qualification and the follow-up that depends on it.

With AI

  • Reads the tender document, or the email body and the thread around it.
  • Extracts the customer, requirement, scope, dates, expected value and contacts.
  • Matches existing accounts and contacts in Oracle Fusion Sales.
  • Creates a draft opportunity for the team to review, confirm and move on.

What it changes

  • Faster turnaround from tender or email to opportunity.
  • No manual re-keying of enquiry details.
  • Fewer missed enquiries and a more complete pipeline.
  • Cleaner, more consistent CRM records.
  • Qualification and follow-up start earlier.
  • More time selling, less time updating the CRM.

Tender Analysis Agent

The challenge

  • Opportunities start with basic details, and the RFPs and requirements arrive later.
  • The business development team reviews documents and updates records by hand.
  • The work is slow and prone to missed or out-of-date information.

With AI

  • Reads tender documents, RFPs and contract information and extracts the key details.
  • Finds the relevant opportunity and compares what it holds with the new documents to spot changes and gaps.
  • Shows the changes for the team to review and approve.
  • Updates the opportunity only with the approved information.

What it changes

  • Less manual effort reviewing long tender documents.
  • Faster tender evaluation.
  • Better data accuracy, because gaps between records and new documents are flagged.
  • Records stay aligned with the latest tender documents.
  • A person approves every update.
  • More time for tender strategy, risk assessment and decisions.

Tender Command Centre

The challenge

  • It is hard to pinpoint why tenders are won or lost.
  • Gaps in the tendering process stay unclear.
  • Lessons that are not captured get repeated.

With AI

  • Analyses the full journey of a tender, from creation to outcome.
  • Identifies gaps, missed requirements and risks.
  • Captures what to retain, avoid and improve.
  • Recommends the key steps for future submissions and applies past insight to upcoming opportunities.

What it changes

  • Stronger submissions built on past insight.
  • Fewer repeat mistakes.
  • Strategy and planning based on data.
  • Lessons that can be reused across future tenders.
  • A clearer view of where preparation can improve.
  • Continuous learning in the tendering approach.

Contract Analyser Agent

The challenge

  • Contract managers assess the risk of each opportunity by hand.
  • The repetitive effort grows with every new opportunity.
  • Without past project insight, assessments vary.

With AI

  • Analyses past risk assessments for patterns.
  • Finds similar historical projects to compare.
  • Recommends the risks, a Red, Amber or Green rating and a Risky (yes or no) flag.
  • Sends the recommendation to the Contract Manager for approval.

What it changes

  • Less effort on each risk assessment.
  • A consistent way to classify risk.
  • Lessons from similar projects flag risks early.
  • Better visibility of high-risk opportunities.
  • Faster review and submission.
  • Better decisions from historical risk patterns.

AI Event Discovery Agent

The challenge

  • Teams search for upcoming events by hand, city by city.
  • Big events are often spotted late, after rooms and venues start filling up.
  • There is no single, current view of events across all the hotel locations.

With AI

  • Finds upcoming conferences, exhibitions, festivals, concerts and weddings in every hotel city, 90 days ahead.
  • Captures the dates, venue, organiser and expected attendance.
  • Ranks events by the hotel business they could bring.
  • Adds them to the CRM, ready to turn into opportunities.

What it changes

  • Earlier visibility of demand in every hotel city.
  • Less manual searching for sales and revenue teams.
  • One event view across all hotels.
  • Smarter pricing and room blocks for peak dates.
  • The top events first, so teams know where to focus.
  • More group and event business captured early.

Customer and account insight

One view of each account, each guest and each team's engagement, with the next step suggested.

Customer 360 Intelligence

The challenge

  • Account details are spread across many reports and screens.
  • Key accounts with falling revenue are hard to spot early.
  • Stalled deals, expiring agreements and relationship gaps go unnoticed.

With AI

  • Shows the account tier, health, revenue trend and year-on-year growth.
  • Shows open leads and opportunities, pipeline value and what has stalled.
  • Shows engagement and stakeholder coverage, and flags expiring records and agreements and overdue tasks.
  • Suggests two or three next steps.
  • In hospitality the same idea gives the front desk a complete guest view from one search: it recognises the guest, personalises the stay, recommends offers and flags open requests.

What it changes

  • Revenue risks spotted early on key accounts.
  • Clear priorities on which deals to chase.
  • Stronger relationships with the right stakeholders.
  • Next best actions suggested by AI.
  • Consistent account reviews across teams.
  • Everything in one view, with no report hunting. It enhances the CRM without replacing it.

Smart Engagements Dashboard

The challenge

  • Engagement data sits across CRM screens, with no single view of physical and virtual meetings.
  • Comparing zones and regions means manual filters and separate reports.
  • Managers spend time finding the story in the data instead of acting on it.

With AI

  • One dashboard for every physical meeting, virtual meeting and customer activity, by zone and region.
  • Anyone can ask in plain language, for example 'Which key accounts had no meeting in 30 days?'
  • A one-click AI summary explains what is driving activity and which zones are rising or slipping.

What it changes

  • One view of all engagement, physical, virtual and more.
  • Answers on demand through plain-language chat.
  • Early warning on slowdowns before they affect results.
  • Zone and region comparison shows where to focus.
  • Faster review and planning with summaries ready to use.
  • More time with customers and less time in dashboards.

Guest Intelligence

The challenge

  • The daily arrival table has dozens of columns for every guest.
  • Teams search row by row for VIPs, loyalty members and complaints.
  • Key guest needs surface late, so service recovery and upsell are missed.

With AI

  • Pulls each hotel's arrival report from OPERA over SOAP.
  • Counts and groups guests into clear modules: VIPs, loyalty, service recovery, extended stays and revenue opportunities.
  • A language model writes an insight and a suggested next action for each module.
  • Flags service recovery, high spenders, missing contacts and birthdays.

What it changes

  • Briefings ready for every hotel each morning.
  • Less searching through columns and guest rows.
  • Focus on the VIP and loyalty members who matter most.
  • Early recovery for guests with past service issues.
  • Revenue opportunities from past usage and long stays.
  • Every insight carries a suggested next step.

Built on the InsightAIQ framework.

See InsightAIQ

Service and CRM data

Agents that classify service emails, suggest the right team and keep contact data clean.

Email Triage Agent

The challenge

  • Several support mailboxes feed many business units.
  • People read each email, find the customer, pick the unit and raise the request by hand.
  • Misrouted emails wait in the wrong queue, and legal complaints can sit among routine ones.

With AI

  • Finds the loan number, ID, mobile number, vehicle number or email in the message, and ignores look-alikes.
  • Matches the customer and loan, then the business unit, with backup checks.
  • The unit's own agent sets the category, severity, urgency, language and a summary.
  • Unclear cases go to a person as an unidentified business unit.

What it changes

  • Faster response, because requests reach the right team at once.
  • Fewer misroutes, because emails are matched to loan records, not guessed.
  • Risk caught early: legal complaints are flagged on arrival.
  • People spend their time on exceptions, with less routine sorting.
  • Consistency: the same rules across every business unit.
  • Understands mixed-language emails, including local languages typed in English letters.

Service Request Agents

The challenge

  • Every food and beverage request goes to a central team first, who pick the right team by hand.
  • Customers describe problems in their own words, with spelling and grammar errors.
  • Agents have no dashboard on their own service request data.

With AI

  • Reads each new request against the assignment policies and suggests the right team on an AI card.
  • Rewrites the customer's problem as clear, well-structured text.
  • A dashboard shows workload, stalled and ageing tickets, turnaround time and risky requests.

What it changes

  • Faster assignment, with less reading and deciding.
  • Fewer wrong handoffs, because suggestions follow policy.
  • Quicker understanding: agents read clear text.
  • Workload balance: see who handles what.
  • Problems caught early: stalled and risky requests show.
  • Clear turnaround, with weekly and monthly trends.

Duplicate Contact Merge Agent

The challenge

  • Duplicate contacts found in the CRM need extra validation.
  • Checking whether a matching member profile exists in the loyalty platform is manual work.
  • Merge requests must be processed only after the required profile checks are done.

With AI

  • Retrieves the duplicate contact records identified in the CRM.
  • Checks Gravty for a corresponding member profile and validates the name match.
  • Starts CRM merge requests for the matching contacts.
  • Updates the non-matching records with placeholder email and mobile values.

What it changes

  • Less manual checking across the CRM and Gravty.
  • Automated validation before any merge request.
  • Merges only after the profile check passes.
  • Less dependence on manual intervention.
  • Better contact data quality.
  • Consistent rules for handling duplicates.

Works across Oracle Fusion Sales and Gravty, the loyalty platform.

Finance and ERP

Agents that turn a file or a request into created, checked and reconciled Oracle Fusion ERP transactions.

Doctor Payout Payables Agent

The challenge

  • A healthcare group receives doctor payout files from each of its centres.
  • Inpatient (IP) and outpatient (OP) payouts are calculated by hand for every doctor.
  • A File-Based Data Import (FBDI) sheet is prepared by hand and loaded into Oracle Fusion ERP, and load errors and reconciliation are checked manually.

With AI

  • The centre's payout file is uploaded in Ask Oracle, with no preparation.
  • The agent reads the file, works out its structure and identifies each doctor and service.
  • It calculates the inpatient (IP) and outpatient (OP) payout for every doctor.
  • It creates the payout invoices in bulk in Oracle Fusion ERP Payables.
  • It checks the result and shows any errors, with the reason for each failure.
  • It matches the invoices back to the original sheet and highlights any differences.

What it changes

  • Faster payouts: invoices for many doctors are created in one run.
  • Accurate IP and OP payouts, calculated the same way every time.
  • Errors visible at once, with the reason for each.
  • Less manual effort: no manual calculation or FBDI preparation.
  • Built-in reconciliation: the invoices are tallied to the original sheet.
  • Scales to many doctors and centres in a single upload.

Runs through Ask Oracle and creates the invoices in Oracle Fusion ERP Payables.

Talk to us about AI on Oracle Fusion at Oracle AI World

Meet the team at Oracle AI World, or book a time that suits you.