PebblePebble
Product
PricingCustomers
Pebble + Front

Agentic support for your hardest Front conversations

Pebble resolves the 30% of Front conversations that need real investigation, not a knowledge-base answer.

Pebble digs through your logs, database, and code, then posts the root cause as an internal comment in Front before your team even opens the conversation. It even creates the customer reply as a shared draft, ready for a teammate to review and send.

Front
A Front conversation with Pebble's shared draft ready for a teammate to sendFrontInboxOpenDraftsSentLaterShared inboxesCustomer SupportSupportSupport · Priority[Slack] SupportCustomer SupportCSV exports failing with a 500 errorpebbleOpenAssigneeJCDDana Larsson · Northgate· just nowOur CSV exports have been failing since this morning with a 500 error —this is urgent, we have a board report due.PebbleInternal commentInternal · not visible to customerRoot causeThe 9:14am deploy changed export date handling — custom date ranges hitan unhandled null. ~40 accounts affected. Fix in ENG-241.@Sam Rivera (Engineering) — can you confirm the fix covers the custom-range case?Shared draftEditable by teammates in Customer SupportPebble edited just nowTake overPebbledrafted a replyHi Dana — sorry for the disruption. We traced it to a change we shipped thismorning and a fix is rolling out now. I'll confirm the moment your exports are back.Nothing has gone to the customer. Your teammate reviews the draft and sends it.

It reads the thread, then investigates the systems where the answer actually lives

When Pebble picks up a conversation it reads the thread, then investigates the systems where the answers live: your codebase, Sentry, Datadog, GCP and AWS logs, a read-only database connection, past conversations, and Slack history.

How it works in three steps

Step 1. Conversation comes in from your customer. Pebble gets to work, investigating across your systems.

A new Front conversation, with Pebble already investigatingFrontInboxOpenDraftsSentLaterShared inboxesCustomer SupportSupportSupport · Priority[Slack] SupportCustomer SupportCSV exports failing with a 500 errorpebbleOpenAssigneeJCDDana Larsson · Northgate· just nowOur CSV exports have been failing since this morning with a 500 error —this is urgent, we have a board report due.Pebble is investigatingInternalCorrelating the 9:14am deploy… reading export logs… checking custom date ranges…CodebaseSentryLogs…DatabasePast conversationsNo teammate has opened the conversation yet.CommentReplyAdd an internal comment…

Nothing goes to a customer until a teammate sends it

Pebble's customer-facing reply is always a Front shared draft: written and ready, but sent by a person. Your teammate reviews, edits, and hits send.

Made for Front's multi-team resolutions

Complex cases need collaboration from more than one team. Pebble puts the technical findings right in the thread they all share.

Front is where support, engineering, finance and the rest of the business resolve hard cases together. Pebble works right in the same thread alongside your team. It posts the root cause as a comment every team can see, so whoever weighs in starts from the same shared context. When a case needs another team, Pebble loops in the right teammate itself by @mentioning them in the conversation, and folds their responses back into the investigation and customer draft automatically.

How is Pebble different from Front's AI?

Front's AI answers from your knowledge base. Pebble investigates your code and logs.

Front's AI handles what your knowledge base already covers: suggested replies for your team, and AI Answers resolving FAQs in chat. Pebble works the conversations those can't close: it reads your code, logs, and database, then posts what it finds for your team to approve. The two coexist, and Pebble works with or without Front's AI. Point your technical conversations at Pebble and it takes them from there.

You decide which conversations Pebble touches

Pebble does not barge into every conversation, you decide how it engages

Tag a conversation.Add the configurable Pebble tag and Pebble goes to work. A Front rule can add that tag automatically, so Pebble starts diagnosing on its own.
Mention @Pebble.An @Pebble in an internal comment pulls it into any conversation, even one it hasn't touched.
It follows the reply.On a conversation it's already tracking, a new customer reply brings Pebble back to fold in the new context and update its comment or draft.
Auto-triage is there if you want it.Optional: Pebble investigates new conversations automatically. Off by default, and most teams scope with a tag instead.
Pebble's settings in Front: trigger tag, a dedicated Pebble teammate, and a response mode with only shared-draft-for-approval or internal-comments-only — no auto-sendPebble settingsCustomer Support inboxFront connectedTrigger tagThe tag that hands a conversation to Pebble. A Front rule can add it automatically.pebbleEditablePebble teammateThe teammate Pebble posts as, so its internal comments attribute cleanly.PebbleDedicated teammate · Customer SupportResponse modeShared draft for approvalA teammate reviews and sends. Default.Internal comments onlyPebble drafts nothing customer-facing.No auto-send — a teammate always sends.

Setup: connect Front, connect your sources, tag a conversation

  1. Connect Front from inside Pebble (Settings → Integrations → Front): a single OAuth sign-in, no subdomain to enter
  2. Optionally create a dedicated “Pebble” teammate for clean comment attribution, and pick your trigger tag
  3. Connect your investigation sources: GitHub for code, Sentry, Datadog, CloudWatch or GCP for logs, a read-only database, your issue tracker. Native integrations for common tools, MCP for the rest.
  4. Tag your first conversation

Tag one conversation to see what it does. The more sources connected, the deeper the investigations.

96%
issue close rate
1 in 3
technical tickets resolved without a code change
< 5 min
to root cause
< 25 min
to a fix ready to review

Common questions

Have more specific questions or want to see how it works?

Let's chat.