FrontLine for Training — multi-client ramps, nesting, and certifications on one engine.
Stand up a 50-agent fintech ramp and a 2,000-agent retail account on the same plans, modules, and nesting engine. Auto-assign training the moment onboarding marks complete. Close the loop from QA finding to refresher in hours, not weeks.
Training dashboard
Acme BPO · 247 agents · 4 clients · 18 plans live
Active plans
18
Assignments in flight
247
Nesting grads · 30d
34
Certs expiring < 30d
6
Active ramps · this week
RetailCo · L1 onboarding ramp · Week 2 of 4
52%
FinCo · Regulated L2 ramp · Week 3 of 6
71%
TelcoPro · French support · Week 1 of 3
18%
AutoCo · CSR re-cert · Continuing-ed
88%
Recent activity
Mike T. graduated nesting — RetailCo L1 · Supervisor sign-off · QA 86 · 14 days
5m
Auto-assigned: FinCo L2 ramp to Sarah K. · ONBOARDING_COMPLETE trigger · plan v3
11m
QA → Coaching → Training: refund-protocol refresher · Eval #4821 · action #2298 · module M-117
32m
Cert expiring: PCI DSS · 6 agents · < 14d window · AutoCo · auto-prescribed re-cert path
1h
Active nesting cohorts
42 trainees
Cohort 47 · RetailCo
Day 8
12 traineesCohort 48 · FinTech
Day 14
8 traineesCohort 49 · MedTech
Day 3
16 traineesCohort 50 · Atlantic
Day 21
6 traineesWhat's hard about training at a multi-client BPO
Generic LMS tools (Cornerstone, Docebo, SAP SuccessFactors Learning) were built for one company, one role catalog, one set of compliance certs. Multi-client BPO breaks every one of those assumptions.
Each client wants its own ramp
RetailCo: 14 days, 5 product modules. FinCo: 21 days, 9 regulatory modules. TelcoPro: 7 days for L1, 28 for L2. A single LMS plan can't reflect that — and a spreadsheet per client is how you lose track of completions.
Did the agent actually pass nesting?
Floor lead says yes, QA says no, the supervisor stopped tracking after week 2. When criteria live in someone's head, the wrong call comes back as a labour-relations file two months later.
New hire just finished onboarding — now what?
Onboarding marks complete; training enrollment becomes someone's manual to-do. The agent waits a day, sometimes three. At 200 hires a year, that's a full FTE burned on a handoff between two systems.
Refresher debt is invisible
QA flags a missed protocol. The coaching follow-up lives in a supervisor's notepad or a three-Slacks-ago message. The same miss surfaces on the next evaluation like a new finding.
Continuing-ed expires; nobody notices
Mandatory training, certifications, recertifications — each decays on its own schedule. Tracked in HRIS or spreadsheets, expiries surface during an audit. By then the contract is at risk.
Classroom session = whiteboard + spreadsheet
Roster, attendance, late, no-show — or it's a notebook on a trainer's desk. When a client asks 'who was in Tuesday's session,' the answer is 20 minutes of forensics in a SharePoint folder nobody owns.
What FrontLine does for your training team
Five training surfaces designed for multi-client operations from the ground up — the plan catalog, auto-assignment from onboarding, nesting with criteria, coaching sessions, and the QA-to-training loop they all feed.
Feature 01
Multi-client plan + module catalog — per-client and per-LOB scoping
One catalog. N clients. Per-LOB modules. Different ramps for RetailCo and FinCo without a spreadsheet per account. Attach, detach, reorder modules on a plan from the UI; partial unique constraints stop the same module from being double-attached.
- Per-client + LOB scoping. Plans carry client_account_id and client_lob_id; modules attach to plans with sequence and required/optional flags. The same engine handles a 50-agent fintech ramp and a 2,000-agent retail account.
- Module library. Reusable modules across plans. Update content once, propagate to every plan that uses it. Versioning preserves audit history of what an agent actually completed.
- Prerequisites + unlock sequence. Module A must complete before Module B. The unlock_after_sequence flag enforces order. Agents see the next module only when the prior one is complete.
- Progress recalculation. Module completions auto-update the parent plan's progress_pct. Trainers see live progress across cohorts, not stale spreadsheet snapshots.
- Audit-grade record. Every plan/module/assignment change is captured: who, when, what. Pull a year of compliance evidence for a client audit in one query.
Plan catalog
All plans
Scoped per client and per LOB · click to drill
RetailCo · L1 onboarding ramp
FinCo · Regulated L2 ramp
TelcoPro · French support N1
AutoCo · CSR continuing-ed
Shared · CASL + privacy refresher
FinCo · KYC refresher (new product)
Per-LOB scoping · auto-assign templates · prerequisites enforced
Feature 02
Auto-assign the moment onboarding completes
When the recruiting → onboarding flow marks an employee complete, the matching training plan auto-assigns. No manual enrolment step. No 2-day handoff gap. The trainer's queue updates in seconds, not in someone-else's-priority.
- ONBOARDING_COMPLETE trigger. The training auto-assign dispatcher fires on the onboarding completion event. The plan-template engine matches by client + LOB + start-date scope and creates the assignment row.
- Duplicate-safe. A partial unique index prevents the same employee from being assigned to the same plan twice. Idempotent — replays don't double-enrol.
- Template-driven. Training plan templates define the auto-assign rules per client/LOB. Adjust the template, every future hire follows the new ramp. No code change.
- Backed by tests. Integration coverage — onboarding /complete and /auto-complete both wired into the auto-assign helper. The next hire from each path lands on the right plan.
Onboarding → Training auto-assign
Onboarding event
ONBOARDING_COMPLETE
Today · 09:14
Sarah K. — FinCo · Compliance · Start: 2026-05-19
Dispatcher fires
Training auto-assignment
Template match: FinCo · Compliance · L2 ramp v3
Plan assigned: FinCo · Regulated L2 ramp
22 modules · 21 days · Day-1 module unlocked
Safety guards
Partial unique index — no double-enrol on replay
Idempotent — failures retry without side-effects
Audit row logged: who, when, which rule, which plan
Feature 03
Nesting that's actually defensible
Per-LOB nesting periods with criteria: minimum days, minimum QA score, supervisor sign-off. The graduation engine evaluates against the configured criteria; ACTIVE → EXTENDED → GRADUATED → FAILED state transitions are auditable. No more 'I think they're ready' graduations.
- Per-LOB criteria. Each LOB defines its own minimum days, minimum QA score, and required supervisor sign-off. RetailCo nests 10 days at 80; FinCo nests 21 days at 85 with a regulated-call sign-off.
- State machine. ACTIVE on assignment, EXTENDED when criteria not met by day-N, GRADUATED on pass, FAILED on hard miss. The state transition is the audit anchor.
- One-active rule. A partial unique index on (tenant, employee, LOB) enforces one active nesting per agent per LOB. No accidental parallel nestings.
- Auditable sign-off. Supervisor signs off in-app with their name, time, and reason captured. The graduation is forensic-grade — defensible at the QBR and at the labour-relations table.
Nesting graduation — Mike T. · RetailCo L1
Mike T. · agent #A-2298
Nesting LOB: RetailCo L1 · started 2026-05-05 · day 14 of 10 min
Criteria
Minimum days
≥ 1014✓Minimum QA score
≥ 8086✓Supervisor sign-off
RequiredSigned✓Supervisor sign-off
P. Singh · 2026-05-19 09:42 · "Ready for floor. Strong on disclosures, mid on objection-handling."
State transition ACTIVE → GRADUATED captured · audit row #au-7821
Feature 04
Coaching sessions with action items and acknowledgement
Coaches build sessions with action items, evidence, and outcomes. Action items route to the agent for acknowledgement. The full chain — session → action → ack → effective — is one record, not a chain of emails.
- Session lifecycle. Schedule, conduct, log outcomes. Each session captures the coach, agent, date, summary, and outcome.
- Action items. Specific, measurable, with deadlines. The agent sees their action items in their portal. The supervisor sees what's overdue.
- Acknowledgement workflow. Agent confirms they've reviewed each action item. The acknowledgement timestamp is the audit anchor — when HR asks 'did they see this coaching,' the answer is the record.
- Manager visibility. Skip-level managers see coaching activity for their org without re-asking the coach. Builds the 'no surprises' culture HR teams want.
Coaching session — Mike T.
Mike T. · RetailCo L1
Session held 2026-05-18 · coach: J. Rivera
Objection-handling deep dive
Action items
Complete module M-117 — Refund protocol refresher
by 2026-05-22
Shadow 2 calls with senior agent — disclosure section
by 2026-05-23
Re-evaluate with sampled call — closing rubric
by 2026-05-26
Agent acknowledged 2026-05-18 14:08 · audit anchor captured
Feature 05
QA finding → coaching action → training module
When a QA evaluator flags a gap, the recommended training module attaches as a coaching action item. The agent's training queue updates. The completion ties back to the original QA evaluation. The loop closes in data, not in someone's memory.
- QA-driven recommendations. QA evaluators flag gaps via the qa_coaching_recommendations table; coaching consumes those recommendations and turns them into action items.
- Action item → module assignment. The action item carries a training_module_id; assigning the action puts the module on the agent's training queue automatically.
- Completion bridge. When the agent completes the module, the action item flips to 'effective.' The QA evaluator sees that the coaching landed — without chasing the supervisor.
- Audit chain QA → Coaching → Training. From the first failing evaluation to the completed refresher, one record. Labour-relations and client-audit defence baked into the workflow.
QA → Coaching → Training
QA evaluation #4821
Missed disclosure on regulated call
RetailCo · evaluator: T. Park · 2026-05-15
Score
68 / 100
Coaching recommendation
Action item #2298 — refresher prescribed
Coaching session c-4421 · due 2026-05-22
Training module
M-117 · Refund protocol refresher
Auto-attached from coaching action · 18 min · score: pass/fail
Completion bridges back: action item → "effective" · QA evaluator notified
The performance loop — QA finding becomes the next training module
QA evaluators flag a coaching recommendation on an evaluation. The coaching team turns it into a session with action items. Each action item carries a training module. The agent completes the module; the action item flips to effective; the QA evaluator sees the loop closed. Same record from the failing call to the proven recovery. The handoff that used to happen in email now happens in data.
- Single audit chain. QA evaluation → coaching recommendation → coaching session → action item → training assignment → completion → effective ack. One record across four teams.
- No emails, no Slack threads. The handoff is in-app. QA doesn't email coaching. Coaching doesn't Slack training. The data does the routing.
- Closure is measurable. The 'effective' state requires the agent to complete the recommended module AND acknowledge the coaching. Not a checkmark — a verified state.
- Manager + skip-level visibility. Both the agent's manager and their skip-level see the chain. Builds the no-surprises culture HR teams want without flooding inboxes.
Performance loop · Mike T.
Mike T. · RetailCo L1
Chain spans 4 teams · one record
QA finding
✓Eval #4821 · missed disclosure · 68/100
Coaching session
✓c-4421 · J. Rivera · 2026-05-18 · agent acked
Action item
✓#2298 · refresher prescribed · due 2026-05-22
Training module
✓M-117 · Refund protocol refresher · completed 2026-05-20
Action: effective
✓Auto-flip · QA evaluator notified · audit row #au-7902
From failing call to verified recovery — one record, four teams, zero handoff emails.
Business outcomes for the people running the BPO
What this looks like in margin terms for the COO or owner.
Ramp compresses — onboarding-to-productive shrinks
Auto-assign on onboarding completion removes the manual enrolment gap. At your hiring volume that compounds into a meaningful capacity recovery without changing your training headcount.
QA miss recurrence drops when the loop closes in data
Closed-loop QA → Coaching → Training assignments reduce repeat misses for the same protocol. The agents with 'effective' action items rarely show the same gap on the next evaluation.
Every graduation defensible at the QBR
Per-LOB criteria + supervisor sign-off + state-machine transitions = a forensic-grade nesting record. When a client questions an agent's readiness, the answer is one screen.
One training team handles N clients without dilution
Per-client/LOB plan templates and auto-assign mean adding a new client doesn't double your training headcount. The team scales linearly with seats, not exponentially with clients.
What training actually gets back
Directional outcomes — real magnitude depends on your prior ramp length, hiring volume, and existing tool stack.
Days faster to nesting
Auto-assign on onboarding completion removes manual enrolment latency. Magnitude depends on your prior handoff gap.
Three systems → one platform
LMS + coaching tracker + nesting spreadsheet collapse into one platform with one audit chain.
Nesting decisions on the record
Per-LOB criteria + supervisor sign-off + state-machine transitions = forensic-grade record of every nesting decision.
Zero lost coaching items
Action items live in data, not on a supervisor's notepad. Acknowledgement is the audit anchor.
Regulatory posture
FrontLine for Training vs. a generic LMS
Six things BPO training teams notice in their first month away from Cornerstone / Docebo / SuccessFactors Learning.
| FrontLine for Training | Generic LMS | |
|---|---|---|
Plans + modules scoped per client + LOB at the data layer Generic LMS treats your workforce as one company. Per-client/LOB scoping is the difference between honest reporting and a spreadsheet per account. | ||
Auto-assign on onboarding completion (event-driven) Some LMS support 'hire = enrol,' few wire it as an event from a separate onboarding system with idempotent guards. | ||
Per-LOB nesting with criteria and supervisor sign-off Nesting is BPO-specific. Generic LMS has no concept of it; you track nesting in spreadsheets or a custom field nobody trusts. | ||
QA → Coaching → Training loop in one record QA in NICE, coaching in a separate tool, training in the LMS. Three systems, three handoffs, no closed loop. | ||
Bilingual platform UI (EN ↔ FR) for agents and trainers Training UI surfaces are fully bilingual today. Course bodies are authored in the language your team writes in; per-course EN ↔ FR sibling-row pairing is on the roadmap. | ||
Audit chain for compliance (PIPEDA, SOC 2) Generic LMS logs course completion but not the chain of changes to course content, plans, or assignments. |
In production today
Training and Coaching (Module 31) is shipped — landed 2026-05-13 with plan/module CRUD, onboarding auto-assign, per-LOB nesting, coaching sessions, and the QA → coaching wire. See the technical detail and feature-level status on the Atlas.
Questions training managers actually ask before signing
Pulled from real fit calls. Short, direct answers.
We already use Cornerstone / Docebo. What does FrontLine add?+
Can we migrate existing courses?+
How exactly does nesting graduation work? We don't want auto-graduation on time alone.+
What if QA finds a gap that doesn't need formal training — just a quick coaching note?+
Can agents see their own training plan and progress?+
Does this replace WFM scheduling for training time?+
Ready to see it in your environment?
Two ways to take the next step.
Other teams using FrontLine
Training doesn't live alone. Here's what the rest of FrontLine looks like from the team next door.