Next week's agenda has three open conflicts. Can you optimise it?
AGENT · SCHEDULE OPTIMIZER
Complex calendars get optimised with the customer's real constraints.
Schedule Optimizer reads the customer's complex calendars (multi-stakeholder meetings, operational slots, clinical agendas), applies the declared constraints (resource availability, priority rules, external deadlines), proposes the consistent optimisation. The agenda owner decides; the agent lays out.
Schedule Optimizer at work.
Conflicts detected: Tuesday 14:00 (board vs strategic customer call), Thursday 15:00 (quarterly close vs product call), Friday morning (two internal meetings with overlapping participants). Proposal: board → Wednesday 09:00; quarterly close stays on Thursday (deadline constraint); product call → Friday 10:00.
Confirmed. Go ahead.
Calendar updated. Notifications sent to participants. Changes are in the audit registry.
Why it exists.
Optimising complex calendars is one of the most time-consuming operational tasks. An operating room with five surgeries a day, a C-level agenda with multi-stakeholder meetings, the operational shifts of a field team — all cases where the constraints to balance are many — resource availability, urgency priorities, external deadlines, participant conflicts — and manual optimisation takes hours a day.
How it works each day.
Schedule Optimizer reads the starting calendar, applies the customer's declarative constraint rules, proposes the structured optimisation. For standard cases it proceeds with the manager's confirmation; for cases with unsolvable conflicts or significant trade-offs, it flags the best option with the detail of the gaps. The agent does not modify the agenda on its own without confirmation.
The decision stays with the manager.
The final decision stays with the agenda owner. The agent does not modify the agenda on its own without confirmation. For changes with significant trade-offs (moving a meeting with an external customer, conflict with a non-negotiable deadline), it proposes the options and waits for authorisation.
From agenda owner to operations director.
Agenda owner
The executive assistant, OR scheduler, shift planner, executive secretariat reclaim manual optimisation time. Capacity concentrates on the final decision and trade-off management, not on the repetitive work of comparing slots and constraints.
fnol.receive 09:14:22 ALLOW
triage.classify 09:14:25 ALLOW
idd.check 09:14:31 WARN
liquidation.propose 09:15:02 ALLOW
SELECT * FROM audit_log WHERE claim_id = '2024-0847'
Downstream operational manager
The head of operating room, head of field team, head of sales see resource utilisation improved. Saturation of constrained resources — operating rooms, C-level windows, specialist staff shifts — increases without increasing workload.
fnol.receive 09:14:22 ALLOW
triage.classify 09:14:25 ALLOW
idd.check 09:14:31 WARN
liquidation.propose 09:15:02 ALLOW
SELECT * FROM audit_log WHERE claim_id = '2024-0847'
Cross-industry operations leadership
The agent is cross-industry with specific verticalisations for healthcare (see OR Scheduling) and for logistics and manufacturing, where operational planning sits at the centre of the flow. Operations leadership gets structured visibility on occupancy and bottlenecks.
fnol.receive 09:14:22 ALLOW
triage.classify 09:14:25 ALLOW
idd.check 09:14:31 WARN
liquidation.propose 09:15:02 ALLOW
SELECT * FROM audit_log WHERE claim_id = '2024-0847'
A weekly agenda with thirty meetings.
Schedule Optimizer reads next week's agenda.
For the head of a mid-market company with a weekly agenda of 30-40 multi-stakeholder meetings, the agent is integrated with the corporate calendar via a dedicated adapter during delivery. On Friday evening it reads next week's agenda on Slack: 34 meetings, 12 recurring participants, 4 protected windows.
Two conflicts identified, structured proposal in the channel.
The agent identifies two conflicts: Tuesday afternoon has the board meeting (5 internal participants) scheduled at the same time as a call with a strategic external customer (3 participants); Thursday afternoon, the quarterly close meeting overlaps with a product call with the main customer. It applies the rules: external customer meetings have priority over internal meetings, quarterly close meetings are protected.
The owner confirms. The agent updates the calendar.
The agent proposes in the channel: move the board meeting to Wednesday morning, keep the Thursday close unchanged, move the product call to Friday morning. The agenda owner reviews the proposal in the Slack thread and confirms. The agent updates the calendar and notifies participants. The event stays in the runtime audit registry.
Declarative rules, versioned constraints.
The Schedule Optimizer rules are declarative. The customer's team defines in a readable format the constraints per calendar type (priorities, protected windows, movement rules), the participants and their declared availability, the non-negotiable external deadlines. The rules live in the customer's repository, versioned, validated at agent startup.
Integration with Google Calendar, Microsoft 365, or the customer's calendar system is delivered via a dedicated adapter during the project by the Exelab team. Technical feasibility depends on the system version and the customer's authorisations.
- Language
- TypeScript (Node.js)
- LLM model
- customer's choice: Anthropic, OpenAI, Mistral, open source models hosted internally, AWS Bedrock for a private model
- Built-in controls used
- pii-detector, topic-guardrail, tool-param-validator
- Native delivery channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- Calendar integration (Google Calendar, Microsoft 365)
- dedicated adapter delivered during the project
- Internal planning system integration
- dedicated adapter delivered during the project
- Memory
- persistent per instance, pgvector + PostgreSQL FTS on customer preferences and patterns
- Registry
- append-only, queryable with a standard SQL client
Frequently asked questions about the agent.
For standard changes pre-approved by the agenda owner (moving an internal meeting within the same week with notification to participants), the agent can proceed on its own. For changes with significant trade-offs (moving a meeting with an external customer, conflict with a non-negotiable deadline), the agent proposes the options; the final decision stays with the owner.
Corporate calendars (Google Calendar, Microsoft 365, Microsoft Exchange) are integrated via a dedicated adapter delivered by the Exelab team during the project. Technical feasibility depends on the system version and the customer's authorisations.
When a calendar presents conflicts unsolvable with the configured constraints (all week slots occupied for a meeting requiring 5 specific participants), the agent flags the impossibility with the detail of the constraints that produced the block. The decision to relax a constraint stays with the owner.
The typical pattern for Schedule Optimizer is 6-10 weeks. Discovery 1-2 weeks, constraint and priority rule configuration 2-3 weeks, corporate calendar integration 2-3 weeks, hand-off to the agenda owner team 1-2 weeks. Effective duration depends on the complexity of constraints and the number of calendar systems involved.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how Schedule Optimizer would configure to the customer's case. Which calendar systems, which priority constraints, which type of agenda to optimise.