Email Service — System Design Academy

Complete System Design Tutorial

The design separates the latency-sensitive online path, the correctness-critical durable transition, and asynchronous projection or repair work. Each chapter below explains one decision shown in the diagrams.

1. Requirements and scope

Functional requirements

Execute the primary email service workflow and return an observable result.
Expose the concrete operations shown below with stable retry identities.
Allow operators to inspect, disable, repair, or replay ambiguous state.
Keep derived analytics and notifications off the correctness-critical path.

Non-functional requirements

Design for 10B messages/day · 50 MB attachments · minutes of retry.
Protect the named business invariant under concurrency and partial failure.
Scale by an explicit aggregate/partition key without one global mutex.
Degrade intentionally when cache, queue, store, or regional ownership is unavailable.

Primary invariant: each accepted email service operation has one stable result and retries do not repeat its business effect.

↑ top

2. Capacity assumptions

State assumptions before choosing technology. The supplied workload is 10B messages/day · 50 MB attachments · minutes of retry. Convert it into peak throughput, retained records, bandwidth, and failure budget; then connect each number to a component.

Dimension	Working estimate	Design implication
Traffic	10B messages/day · 50 MB attachments · minutes of retry	Separate online compute, durable state, and asynchronous consumers.
Peak factor	5–10× average	Admission control, autoscaling headroom, and hot-key protection.
Stored state	write rate × retention × record bytes	Partition count, indexes, compaction, backup, and archival tier.
Network	peak operations × payload size	Batching, compression, edge placement, and egress budget.
Availability	99.9–99.99%	Multi-zone replicas, tested failover, and dependency error budgets.
Recovery	RPO/RTO by invariant	Ack boundary, replay retention, reconciliation, and operator runbooks.

Do not cargo-cult the numbers: show the arithmetic, name the uncertain variable, and say which production metric will replace the interview estimate.

↑ top

3. API and data model

POST /v1/messagessubmit idempotent mail

SMTP RECEIVEaccept external message

GET /v1/mailbox?cursor=sync changes

POST /v1/messages/{id}/actionslabel/delete

Authoritative records and indexes

message(message_id PK, body_uri, content_hash, sender)

delivery(message_id, recipient, domain, attempt, state)

mailbox_entry(user_id, message_id, labels, thread_id, version)

domain_queue(domain, next_attempt, backoff, reputation)

Command and event envelopes

# Command envelope
operation: POST /v1/messages
request_id: req_01J...
idempotency_key: idem_7f2...
aggregate_key: email-service_123
expected_version: 17
payload: {...}

# Durable event
event_id: evt_01J...
aggregate_key: email-service_123
event_type: email_service_committed
schema_version: 3
trace_id: tr_8ac...

The authoritative aggregate and idempotency record share a transactional or conditional-write boundary. Event streams, search indexes, caches, analytics, and dashboards are rebuildable projections.

↑ top

4. Why this architecture and what it costs

The boxes are not a shopping list. Each boundary exists because it owns a different invariant, scaling dimension, data lifecycle, security policy, or failure domain. The table makes those choices explicit and shows when the design should remain simpler.

Component	Owned responsibility	Why separate it	Rejected alternative	Cost and evolution trigger
Mail Client	Owns client protocol, request identity, retry, and resume behavior. In this design it participates in: Authenticate submission and durably store the immutable MIME body.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents the external caller from also owning the contract and failure behavior of Submission API.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Remote domain remains unavailable for days” to the external caller.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Mail Client boundary before extracting or scaling it.
Submission API	Owns authentication, validation, quotas, deadlines, and routing. In this design it participates in: Expand recipients, filter policy/malware, and create independent delivery records.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Mail Client from also owning the contract and failure behavior of SMTP Ingress.	Put protocol, authentication, quotas, and business transitions in one process. Acceptable at low scale, but it couples external traffic risk to domain releases. Here that simpler choice would couple “Recipient expansion creates huge fan-out” to Mail Client.	Costs another hop, policy distribution, and a high-blast-radius fleet. Introduce it when many clients or services need one protocol and protection boundary. Measure the latency, saturation, lag, or mismatch at the Submission API boundary before extracting or scaling it.
SMTP Ingress	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Submission API from also owning the contract and failure behavior of Identity / DKIM.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Spam verdict changes after delivery” to Submission API.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the SMTP Ingress boundary before extracting or scaling it.
Identity / DKIM	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: On local delivery append mailbox entry and change cursor; external delivery records response.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents SMTP Ingress from also owning the contract and failure behavior of Spam / Malware.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Client retries submission after timeout” to SMTP Ingress.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Identity / DKIM boundary before extracting or scaling it.
Spam / Malware	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Process bounces/complaints idempotently and update reputation plus user-visible state.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Identity / DKIM from also owning the contract and failure behavior of Durable Mail Queue.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Mailbox index lags message delivery” to Identity / DKIM.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Spam / Malware boundary before extracting or scaling it.
Durable Mail Queue	Preserves durable ordering, replay, and consumer checkpoints. In this design it participates in: Authenticate submission and durably store the immutable MIME body.	Decouples bursts and failure from user acknowledgment. It prevents Spam / Malware from also owning the contract and failure behavior of Domain Scheduler.	Call every downstream consumer synchronously. That couples user latency and availability to analytics, notification, indexing, and audit dependencies. Here that simpler choice would couple “Remote domain remains unavailable for days” to Spam / Malware.	Costs eventual consistency, schema evolution, replay tooling, idempotent consumers, and lag operations. Earn it when bursts or downstream failure must not block acknowledgment. Measure the latency, saturation, lag, or mismatch at the Durable Mail Queue boundary before extracting or scaling it.
Domain Scheduler	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Expand recipients, filter policy/malware, and create independent delivery records.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Durable Mail Queue from also owning the contract and failure behavior of SMTP Egress.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Recipient expansion creates huge fan-out” to Durable Mail Queue.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Domain Scheduler boundary before extracting or scaling it.
SMTP Egress	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Domain Scheduler from also owning the contract and failure behavior of Message Blob Store.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Spam verdict changes after delivery” to Domain Scheduler.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the SMTP Egress boundary before extracting or scaling it.
Message Blob Store	Owns authoritative versioned state or immutable evidence. In this design it participates in: On local delivery append mailbox entry and change cursor; external delivery records response.	Defines the source of truth and commit boundary. It prevents SMTP Egress from also owning the contract and failure behavior of Mailbox Metadata.	Let each service persist its own interpretation. That removes one clear commit boundary and makes retries, support, and reconciliation ambiguous. Here that simpler choice would couple “Client retries submission after timeout” to SMTP Egress.	Costs schema migration, replication, backup, and shard ownership. Split storage only when access patterns, retention, or compliance boundaries materially differ. Measure the latency, saturation, lag, or mismatch at the Message Blob Store boundary before extracting or scaling it.
Mailbox Metadata	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Process bounces/complaints idempotently and update reputation plus user-visible state.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Message Blob Store from also owning the contract and failure behavior of Thread / Search Index.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Mailbox index lags message delivery” to Message Blob Store.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Mailbox Metadata boundary before extracting or scaling it.
Thread / Search Index	Serves a low-latency, disposable projection with authoritative fallback. In this design it participates in: Authenticate submission and durably store the immutable MIME body.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Mailbox Metadata from also owning the contract and failure behavior of Push / IMAP Sync.	Read authoritative storage for every request. Simpler initially, but latency and source-of-truth load scale together and a hot key becomes a database incident. Here that simpler choice would couple “Remote domain remains unavailable for days” to Mailbox Metadata.	Costs invalidation, staleness, stampede protection, and memory. Add or split it only after the workload (10B messages/day · 50 MB attachments · minutes of retry) proves repeated reads dominate. Measure the latency, saturation, lag, or mismatch at the Thread / Search Index boundary before extracting or scaling it.
Push / IMAP Sync	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Expand recipients, filter policy/malware, and create independent delivery records.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Thread / Search Index from also owning the contract and failure behavior of Bounce Processor.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Recipient expansion creates huge fan-out” to Thread / Search Index.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Push / IMAP Sync boundary before extracting or scaling it.
Bounce Processor	Executes idempotent asynchronous work under lease and retry policy. In this design it participates in: Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Push / IMAP Sync from also owning the contract and failure behavior of Reputation Service.	Perform the work before acknowledging the caller. This gives simpler consistency but unbounded tail latency and fragile retries. Here that simpler choice would couple “Spam verdict changes after delivery” to Push / IMAP Sync.	Costs leases, checkpoints, retries, DLQs, and lag operations. Separate once execution time or burstiness no longer fits the request deadline. Measure the latency, saturation, lag, or mismatch at the Bounce Processor boundary before extracting or scaling it.
Reputation Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: On local delivery append mailbox entry and change cursor; external delivery records response.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Bounce Processor from also owning the contract and failure behavior of Audit / Abuse.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Client retries submission after timeout” to Bounce Processor.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Reputation Service boundary before extracting or scaling it.
Audit / Abuse	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Process bounces/complaints idempotently and update reputation plus user-visible state.	Proves correctness and supports controlled recovery. It prevents Reputation Service from also owning the contract and failure behavior of operators and repair tooling.	Infer correctness from request logs. Logs do not prove durable business state, detect silent drift, or safely repair ambiguous outcomes. Here that simpler choice would couple “Mailbox index lags message delivery” to Reputation Service.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Audit / Abuse boundary before extracting or scaling it.

Decision 01

Chosen boundary vs simpler monolith

Acknowledge only after durable queueing; remote SMTP delivery can retry for days without keeping client connections open.

A single deployable with one transactional store is the right starting point for a small email service workload. The proposed boundary becomes justified when acknowledge only after durable queueing; remote smtp delivery can retry for days without keeping client connections open. Separating it contains the named failure “Remote domain remains unavailable for days” without placing unrelated work on the critical path.

Decision 02

Correctness vs availability

Separate immutable message bodies from mailbox membership and labels so one message can appear in multiple folders without copying content.

The design coordinates only where each accepted email service operation has one stable result and retries do not repeat its business effect. Everywhere else it accepts bounded staleness or asynchronous propagation. The alternative—strong coordination for every projection—simplifies the mental model but increases tail latency and turns partitions into broad write unavailability.

Decision 03

Operational cost and evolution

Treat external domains as slow and adversarial; per-domain queues, reputation, backoff, and bounce state contain failures.

This choice adds dashboards, ownership, schema compatibility, capacity planning, and recovery runbooks. Start with the minimum topology that supports “Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.” and extract another service only when measured throughput, blast radius, data lifecycle, or team ownership changes independently.

Why this design: the synchronous path contains only the decisions required to protect each accepted email service operation has one stable result and retries do not repeat its business effect. Durable projections, notifications, indexing, analytics, and repair are isolated so their load or failure cannot corrupt the acknowledged operation.

Correctness boundary

Acknowledge only after durable queueing; remote SMTP delivery can retry for days without keeping client connections open.

Scale boundary

Separate immutable message bodies from mailbox membership and labels so one message can appear in multiple folders without copying content.

Operational boundary

Treat external domains as slow and adversarial; per-domain queues, reputation, backoff, and bounce state contain failures.

↑ top

5. Critical flow and algorithm

Authenticate submission and durably store the immutable MIME body.

Expand recipients, filter policy/malware, and create independent delivery records.

Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.

On local delivery append mailbox entry and change cursor; external delivery records response.

Process bounces/complaints idempotently and update reputation plus user-visible state.

Retry-safe mutation skeleton

identity = authenticate(request)
prior = idempotency.get(identity, request.key)
if prior: return prior.result

current = authoritative_store.read(partition_key)
next_state, events = domain.transition(current, request)
authoritative_store.compare_and_set(current.version, next_state)
idempotency.commit(request.key, next_state.result)
outbox.append(events)
return next_state.result

Exactly-once business effect: the network can duplicate requests. Stable identities, conditional state transitions, immutable event IDs, and reconciliation make the user-visible effect happen once.

↑ top

6. Caching, hot keys, and backpressure

Cache policy

Cache immutable or slowly changing reads with versioned keys, TTL jitter, short negative entries, and request coalescing.

Hot partition or tenant

Measure per-key and per-tenant skew. Replicate safe reads, isolate the offender, salt only when ordering permits, and enforce budgets.

Backpressure

Bound queues, buffers, connection pools, and concurrency. Shed low-priority work and scale from lag/oldest-item age rather than CPU alone.

Stampede and retry storms

Use single-flight/coalescing, jittered expiry, bounded retries, exponential backoff, and circuit breakers so one dependency failure does not multiply traffic.

Cache outage rule: never allow every miss to fall through simultaneously. Use local protection, request coalescing, circuit breakers, bounded pools, and controlled fallback.

↑ top

7. Partitioning and consistency

partition = stable_hash(primary_or_aggregate_key) → shard owner
write = conditional_update(expected_version)
read_projection = versioned and rebuildable

Strong boundary

each accepted email service operation has one stable result and retries do not repeat its business effect. Use a uniqueness constraint, serializable/conditional write, expected version, or fencing token at this boundary.

Eventual boundary

Analytics, notifications, search, read models, and operational dashboards may lag when freshness is observable and replay exists.

Operation	Consistency	Reason
Authoritative mutation	Conditional / serializable	Protect uniqueness, ownership, money, inventory, order, or legal state transition.
Read after own write	Primary, session token, or cache warm	The caller must observe the result it just committed.
Ordinary read projection	Bounded stale where safe	Replicas and caches improve latency when version/freshness is visible.
Analytics and operations	Eventual + replay	Freshness is less important than isolating the online path.

Route every mutation for one aggregate to one owner/partition epoch.
Use cursors rather than offsets for changing ordered collections.
Retain tombstones until replication and repair windows close.
Move shard ownership through versioned maps and dual-read verification.

↑ top

8. Failure drills and recovery

Failure	Runtime behavior	Recovery proof
Remote domain remains unavailable for days	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Recipient expansion creates huge fan-out	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Spam verdict changes after delivery	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Client retries submission after timeout	Query/retry with the same operation identity; do not issue a second business effect.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Mailbox index lags message delivery	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.

Recovery checkpoint

# Recovery checkpoint
scenario: Remote domain remains unavailable for days
operation_id: op_01J...
aggregate_key: email-service_123
last_durable_version: 18
projection_checkpoint: partition-07@offset-482901
retry_policy: same_identity / bounded_backoff
reconciliation_status: pending
owner: Audit / Abuse

Every dependency needs a timeout, retry budget, circuit policy, and explicit fail-open, fail-closed, stale, or read-only mode. “Highly available” is incomplete until degraded behavior is named.

↑ top

9. Security, regions, and observability

Security and abuse

Authenticate at ingress and authorize the exact resource/tenant on every operation.
Encrypt transport and durable data; tokenize or isolate sensitive material.
Rate-limit by tenant, actor, resource, and network; detect enumeration and anomalous fan-out.
Redact secrets and personal data before logs or analytical storage.
Audit privileged mutations, support access, policy changes, and forced recovery.

Signals that prove correctness

User-visible success rate and p50/p95/p99 latency.
Timeouts, circuit state, saturation, and replica health by dependency.
Queue lag, oldest-item age, retries, and dead-letter growth.
Conflict, duplicate-suppression, and reconciliation mismatch counts.
Business drill: Remote domain remains unavailable for days — alert before the user-visible invariant is violated.

Regional evolution

Begin multi-zone with tested failover, backups, and reconciliation.
Add global stateless ingress and regional read projections.
Keep one write owner per aggregate or partition where coordination is expensive.
Move ownership with epochs/fencing; avoid silent simultaneous writers.
Use expand/contract schemas, canaries, shadow reads, and reversible cutovers.

↑ top

10. Interviewer follow-ups

How does SMTP retry work?

Acknowledge only after durable queueing; remote SMTP delivery can retry for days without keeping client connections open. Walk the concrete lifecycle step “Authenticate submission and durably store the immutable MIME body.” Then test it against “Remote domain remains unavailable for days”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How are threads built?

Separate immutable message bodies from mailbox membership and labels so one message can appear in multiple folders without copying content. Walk the concrete lifecycle step “Expand recipients, filter policy/malware, and create independent delivery records.” Then test it against “Recipient expansion creates huge fan-out”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do attachments deduplicate?

Treat external domains as slow and adversarial; per-domain queues, reputation, backoff, and bounce state contain failures. Walk the concrete lifecycle step “Schedule per-domain SMTP attempts with reputation, quotas, and exponential backoff.” Then test it against “Spam verdict changes after delivery”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How does spam filtering evolve?

Acknowledge only after durable queueing; remote SMTP delivery can retry for days without keeping client connections open. Walk the concrete lifecycle step “On local delivery append mailbox entry and change cursor; external delivery records response.” Then test it against “Client retries submission after timeout”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do clients synchronize offline?

Separate immutable message bodies from mailbox membership and labels so one message can appear in multiple folders without copying content. Walk the concrete lifecycle step “Process bounces/complaints idempotently and update reputation plus user-visible state.” Then test it against “Mailbox index lags message delivery”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

Answer pattern: state the invariant, choose a mechanism, walk one failure, name the recovery signal, and explain the cost of the alternative.

↑ top

11. How to present this in 45 minutes

Time	What to cover
0–5 min	Clarify scope, workload, SLOs, and the correctness invariant.
5–10 min	Estimate throughput/storage and define concrete APIs and records.
10–20 min	Draw online, durable-write, and asynchronous/repair paths.
20–30 min	Explain the exact lifecycle, partition key, and concurrency control.
30–38 min	Discuss caching, hot keys, replication, and backpressure.
38–43 min	Run failure drills, security, observability, and regional evolution.
43–45 min	Summarize the five most important choices and trade-offs.

Final decision summary

Invariant	each accepted email service operation has one stable result and retries do not repeat its business effect.
Partitioning	Stable aggregate/key hash with versioned ownership and measured hot-key isolation.
Commit	Expected version or uniqueness at the authoritative state boundary.
Async work	Durable outbox/log, idempotent consumers, checkpoints, and replay.
Recovery	Same operation identity, durable truth lookup, and business reconciliation.

Closing summary

CLOSING “The design protects each accepted email service operation has one stable result and retries do not repeat its business effect. The online path stays bounded, authoritative state is versioned and partitioned, and asynchronous work is replayable and observable.”

↑ top

Mail Client

Submission API

SMTP Ingress

Identity / DKIM

Complete System Design Tutorial

1. Requirements and scope

Functional requirements

Non-functional requirements

2. Capacity assumptions

3. API and data model

Authoritative records and indexes

Command and event envelopes

4. Why this architecture and what it costs

Chosen boundary vs simpler monolith

Correctness vs availability

Operational cost and evolution

5. Critical flow and algorithm

Retry-safe mutation skeleton

6. Caching, hot keys, and backpressure

7. Partitioning and consistency

Strong boundary

Eventual boundary

8. Failure drills and recovery

Recovery checkpoint

9. Security, regions, and observability

Security and abuse

Signals that prove correctness

Regional evolution

10. Interviewer follow-ups

11. How to present this in 45 minutes

Final decision summary

Closing summary