Retroactive record — Branch Bot Response: Customer/Contact Field
Provenance note for an initiative that was built, GA'd, and largely closed before it was ever recorded in this repo. Written so the reconstruction is auditable: what is confirmed, what was inferred, and what remains open.
Context
A jira-alignment-check on BOT-4308 — an intermittent
500 when saving a bot-response branch with a second condition — found the Bug unparented (child of
no Epic). Tracing its Relates / Problem/Incident links led to Epic
BOT-3864, "26Q2 Branch Bot Response Logic:
Customer/Contact Field": 11 Stories and ~35 Tasks spanning 2026-02-26 → 2026-06-24, already run on
this repo's per-user-story convention — and completely absent from the repo.
The PM then supplied the Confluence PRD
(last modified 2026-04-16), which is the substantive source. The decision (2026-07-27) was to record it
as a 2026-Q2 Chatbot initiative with status: shipped, anchored to the existing Epic BOT-3864, and
to route the unfinished tail to the Engineering Initiatives bucket
rather than reopen the initiative.
Notes
Confirmed (read from a file or a Jira/Confluence response).
- Epic BOT-3864 exists in project BOT,
issuetype: Epic, status Done. - Its 11 Stories: BOT-3928, BOT-4043 – BOT-4052. Six are Done; five are still
To Do— BOT-4044, BOT-4045, BOT-4047, BOT-4048, BOT-4052. - BOT-4308 has no
parentfield and noWork item splitlink. The absence is real, not a field-filter artifact: the same-shaped query returned a populatedparentfor BOT-4303 and BOT-3931. - The PRD's §1–5 graded sections map 1:1 onto those Jira Stories; its 1.1–7.1 table's JIRA Ticket
column is entirely
TEMP-<uuid>placeholders. - PRD roles: PM Dimas Fauzi Hidayat · Designer Wulan Febyazzahra Putri · Engineers Puji Triwibowo, Eko Aprianto · QA Muhammad Izzul · PMM Yosephine Dhisaclara. Confluence "Team" field: Hadiningbot.
- GA date in the PRD launch plan: 2026-04-01, "UI Configuration will follow after that".
Inferred or decided, not read.
progress_pct: 55is story-count-based (6 of 11 Stories Done), not man-days. No effort was ever captured for this initiative, so everyeffort_*field isTBDby design and no forecast exists.qa_lane: Bwas classified on 2026-07-27 from the PRD's content (real-time CDP fetch on the live conversation path; read-only PII handling), not by theqa-lane-advisor.- Composite story ids
BBC-S01…BBC-S18were assigned during this port — they did not exist before and appear in no Jira ticket. They are now stable keys; do not renumber them.
Conflicts left standing (not silently resolved).
- Staffing disagrees with the board. The PRD names Puji Triwibowo and Eko Aprianto as engineers, but the Jira work was assigned to Agus Suparman (BE — BOT-3931, BOT-4308) and Samudra Linggar (FE — BOT-4303). The README records board reality per the 2026-07-27 decision; the PRD's own role table is preserved verbatim in the PRD header. The divergence is unexplained.
- Epic Done vs. five open Stories. BOT-3864 is closed while BOT-4044/4045/4047/4048/4052 remain
To Do. Either the Epic was closed prematurely or those Stories are abandoned; the board does not say which. - Two overlapping story sets in the PRD, merged into one canonical §8 table. Every merge-exposed
conflict is recorded as
OQ-1…OQ-6in the PRD, notably the AND/OR Must-vs-Should disagreement (OQ-1) and numeric fields being specified but absent from the field catalogue (OQ-2).
A defect pattern worth naming. BOT-4308 is not the only unparented branch-response Bug.
BOT-4274 ("Field condition — Select Match can not be
save", Done) carries the identical branch-response + chatbot-UI label pair and is also
NO-PARENT. Separately, four branch-related escalations came through as Fast Track under different
Epics: BOT-4442,
BOT-4302 (Won't Fix),
BOT-3630,
BOT-3626 — all labelled jira_escalated / qon-p2.
QA-filed branch-response defects appear to routinely bypass Epic parenting.
Decisions / changes
- Created
chatbot/branch-bot-response/— README (2026-Q2,shipped, anchor BOT-3864) + the ported PRD + this note. - Roadmap: Portfolio row (Health
Completed), QA Lanes row, and a Portfolio-progress row added. Deliberately not added to## Priority— a shipped prior-quarter initiative carries no PM rank. - Capacity: deliberately not given a numeric allocation.
capacity.mdis scoped to 2026-Q3; this initiative's work landed in Q2, so loading it into the Q3 plan would misstate current headroom. A## Notesline records the omission.
Board writes applied 2026-07-27
Executed on the DRI's go-ahead, each verified by re-reading the issue:
- BOT-4308 → parent BOT-3864 (confirmed). No
Work item splitlink added: a QA-filed defect has no owning PRD user story to split from. - BOT-4274 → parent BOT-3864 (confirmed) — the second unparented branch-response Bug.
- BOT-4044, BOT-4045, BOT-4047, BOT-4048, BOT-4052 → Done (transition id 121; all five confirmed). Closed on the DRI's statement that the work had already deployed. All 11 Stories under BOT-3864 are now Done. This was bookkeeping, not delivery — no per-Story code evidence was gathered, so the resulting 100% rests on attestation.
- Conflict #2 above (Epic Done vs. five open Stories) is therefore resolved; conflict #1 (staffing divergence) and #3 (the merged story sets' open questions) still stand.
Follow-ups
Not applied — listed for the DRI/TPM to action:
- Fix BOT-4308 — the only open work, and it now sits unfixed under a Done Epic. Decide whether it
stays on BOT-3864 or moves to a bucket story under Engineering Initiatives
(BOT-4620); the bucket routing was not executed,
because an
A<n>tracker row is per-story and BOT-4308 is a Bug. - Consider a QA intake check so branch-response Bugs get an Epic at file time — two of them were filed unparented.
- Answer OQ-1 … OQ-5 in the PRD; OQ-2 (numeric fields) is the one that changes buildable scope. Note OQ-1 is now moot for scheduling (BOT-4045 is closed) but still a spec inconsistency.
- Confirm the staffing divergence — was the PRD's engineer roster ever accurate?
- Wire
sources:ifproduct-intelligence-hubis checked out; the PRD currently declaressources: []with an evidence-gap note.