Assignment that knows who's actually available
Most teams keep leave in a separate HR tool the engineering manager never opens while handing out work. Infersync ties leave management to the AI assignment ranker, so when you (or the autopilot) assign a task, it routes around anyone who's out next Wednesday or already at capacity instead of overcommitting them.
Why teams end up here
- Leave lives in a separate HR tool (BambooHR, Personio) that the engineering manager doesn't open when assigning work.
- Someone's leave gets approved on Wednesday; nobody reflects it when work is handed out; the team finds out on Monday standup.
- Work gets assigned on optimistic availability, then slips when leave, sick days, and cross-team support eat the buffer.
- When a manager wants to know 'who's actually free next week', the answer requires opening four tools and doing mental arithmetic.
- The cheapest skilled person gets piled on because nobody can see they're already over-committed.
Leave and availability, in the assignment, no manual reconciliation
Leave requests in Infersync are approved in-app and become part of each member's availability automatically. The AI assignment ranker scores every candidate on skill fit, real availability (working hours minus leave minus already-committed work), and cost, so work routes to whoever can actually take it, and someone who's out gets skipped before the assignment lands. No separate HR tool to check, no spreadsheet to reconcile.
Leave management with availability gating
Leave requests submitted in-app, approved by manager or admin, then folded into each member's availability automatically. No manual sync, no separate HR tool to remember to open.
Leave-aware AI assignment ranking
The assignment ranker scores candidates on skill_fit × availability × cost. Members on leave in the window are excluded from availability before the task is offered, so work never lands on someone who's out.
Real availability, not a guess
Availability is working hours minus approved leave minus already-committed work, computed per member, surfaced as a single number instead of a four-tool spreadsheet calculation.
Clock in / out and break tracking
Daily clock in/out, work-item timers, and break tracking keep the underlying hours honest, so availability and cost stay accurate as the week runs.
Daily clock-in reminders
Auto-pinged at scheduled start time so attendance data stays honest (under-clocked hours throw off availability).
Steps to get from zero to live
- 1
Connect GitHub + invite team
OAuth, invite members, set per-person working hours and weekly availability in workspace settings.
- 2
Members submit leave in-app
Leave request from any page, approved by manager / admin. Approved leave is folded into that member's availability automatically.
- 3
Assign work that respects availability
Hand out work yourself or ask the autopilot to assign it; the ranker skips anyone on leave or already at capacity, so the cheapest skilled person available gets the task.
- 4
Keep the numbers honest
Daily clock in/out, work-item timers, and break tracking keep availability and cost accurate as work runs.
Available on the Operations plan and up
Time tracking and leave management ship on Base (£7/seat/month). The AI assignment ranker that consumes availability and leave lives on Operations (£15/seat/month). The 14-day free trial gives every workspace Operations-tier features so you can see the leave-into-assignment loop on real data.
FAQs about assignment that knows who's actually available
How does the leave-to-availability linkage actually work?
When a leave request is approved (by manager or admin), it's persisted as a date range on the member's record. The AI assignment ranker reads that range and excludes leave days from availability scoring, so a member who's out is ranked below available teammates for work in that window. There's no separate sync step, leave is part of the member's availability record by definition.
Do leave requests need approval, or is it self-serve?
Per workspace setting. Default is 'requires approval from manager or admin'. Self-serve mode (auto-approve for members) is available for teams that prefer trust over process. The approval audit log captures who-approved-what-when for compliance.
What about sick days and unplanned absence?
Sick days go through the same leave system but flagged as 'sick' rather than 'planned'. Availability treats them identically once approved (or marked taken). Unplanned absence on a specific day can be back-dated by manager or admin to keep the data accurate after the fact.
Does this replace BambooHR / Personio for leave?
For engineering-team leave specifically, yes, the in-app flow plus availability link is faster than a separate HR system. For company-wide HR (payroll, benefits, broader employee records), no, keep your HR tool for that and use Infersync for the engineering-team assignment loop. Two-way sync to BambooHR / Personio is on the roadmap, not shipped today.
How does the AI know about availability?
The autopilot assignment ranker uses (skill_fit × availability × cost_inverse). 'Availability' is the member's remaining capacity in the next 7 days, computed from working hours minus leave days minus already-committed work. When you ask the AI to 'assign this to the cheapest skilled person available', it reads that availability directly. Leave changes propagate to the ranker on the next assignment without a manual refresh.