What's new

Release notes for each sprint, most recent first.

Sprint 26 — Phase-In v2 UI Latest

5-zone Phase-In result view — The ECO Phase-In tab has been completely rebuilt. The new view shows: a decision header with synchronised date, driver item, and data confidence score; a financial exposure strip with four metric cards (stock at risk, PO exposure, total risk, MOQ saving); per-item cards with SVG timeline and reasoning bullets; a what-if simulator for testing alternative dates; and a customer delivery commitments table.

What-if simulator — Enter a target date or use quick-shift buttons to test ±4 weeks. See feasibility, financial delta, and stranded items. Save scenarios for comparison and commit when ready, with the option to notify affected assignees.

Sales order impact — Upload customer sales orders via CSV on the ERP Data page. The Phase-In tab shows which customer deliveries are at risk if the phase-in date shifts.

Drift detection banner — When ERP data changes after a calculation, an amber banner appears showing which items shifted and by how many days. One-click recalculate to update.

Warehouse location settings — New settings page at Settings → Warehouse Locations. Toggle which stock location types (standard, quarantine, returns, scrap, transit, consignment) count as available inventory for phase-in calculations.

URGENT action status — Actions that become overdue after a scenario commit are automatically marked URGENT. Urgent actions sort to the top of their department group with a red border accent.

Sprint 25 — Phase-In Calculator v2 Advanced

MOQ analysis — For each PO arriving after the phase-in date, Redline checks whether the quantity can be reduced to the minimum order quantity instead of cancelling. Suggestions include the PO number, current vs suggested quantity, and potential EUR saving.

EC-level synchronisation — Multi-item ECOs now synchronise to a single phase-in date (the latest across all non-phantom items). The driver item is identified and highlighted.

What-if simulation engine — Pure deterministic simulation: given a target date, returns stranded POs, stranded production orders, financial delta, and feasibility assessment (feasible / warning / infeasible). No side effects — save and commit are separate operations.

Drift detection — Background job (every 6 hours) compares stored phase-in dates against fresh calculations. Flags items where the date has shifted, with delta and change drivers.

Scenario management — Save what-if scenarios for later comparison. Commit a scenario to promote it to the actual phase-in date, with proportional recalculation of action due dates.

Sprint 24 — Phase-In Calculator v2 Foundation

Stock location filtering — Phase-in calculation now filters stock by warehouse location type. Only locations marked as “available” in tenant settings count as consumable inventory. Quarantine, returns, and scrap locations are excluded by default.

Where-used analysis — For each affected item, Redline identifies parent assemblies and checks whether they have active production orders or committed sales orders.

Financial exposure calculation — At-risk stock quantity, excluded stock, and financial exposure in EUR calculated per item using standard cost from the item master.

Sales order awareness — The calculation now considers customer delivery commitments when assessing the impact of a date change.

Data confidence scoring — Each item receives a 0–100 confidence score based on data completeness (missing lead times, stale snapshots, unresolved items reduce the score).

Sprint 23 — Revision-aware BOM splitting

Revision integrity on BOM import — When importing a BOM CSV, Redline now detects the parent item’s revision and checks whether it exists in the item master. If the revision is missing, the import is blocked with a clear warning. You can create the revision automatically (copies description from the existing revision) or create it manually first.

Revision-aware item detail — The BOM tab and Where Used tab on item detail pages now filter by revision. Rev A and Rev B of the same item each show their own BOM. Navigation between items preserves the correct revision context.

BOM export improvements — The Export BOM CSV now includes Parent Rev, EC Phase-In, and EC Phase-Out columns. All columns are ordered to match the on-screen table layout.

Import speed — BOM import reduced from approximately 10 seconds to 2-3 seconds by validating in-memory before writing to the database.

UI fixes — The Process button now redirects back to the ERP Data page. The status stepper shows green checkmarks when a BOM is successfully processed. The re-validate flow no longer hides the Process button. Source values display as coloured pill badges. Missing columns (Component Rev, Eff From/To) added to the BOM tab.

Sprint 22 — AI Profile + Action List improvements

AI Profile — New settings page at Settings → AI Profile. Feed in your company description, preferred terminology, ERP system name, writing tone, and worked examples. All AI-generated action instructions, phase-in reasoning, and action suggestions apply this context from the next call onwards. Existing action text is not retroactively updated.

AI suggestion rationale — The Suggest Actions panel on ECR and EC action lists now shows a rationale explaining why those specific actions were recommended, and a per-suggestion reason on each line. Suggestions are deduplicated against existing actions and capped at five per call.

Due date improvements — Due date fields on EC action lists now show a calendar picker icon. The field accepts typed input correctly. A new Set due date for all control above the action list lets coordinators push a single date to all actions at once.

Bug fixes — Priority set on a preliminary action in the ECR now correctly carries over to the EC when the ECR is approved. Previously all carried-over actions defaulted to Medium. The Responsible Engineer column on the Engineering Changes list now resolves external contacts to their full name.

Sprint 21A — BOM Upload Wizard

3-step BOM wizard — BOM upload is now a guided wizard (Upload → Review & Fix → Confirm) on a dedicated page at /engineering/[id]/bom. Nothing writes to the database until you confirm in Step 3. The old inline upload form on the EC detail page has been replaced with summary cards and a link to the wizard.

Instant client-side preview — CSV files are parsed in the browser with immediate column mapping preview, so you can verify the data before sending anything to the server.

Connector abstraction — The new BomSource architecture supports pluggable BOM sources. CSV is the first implementation; PLM API and ERP connectors (Exact, D365) will follow in future sprints.

Validation cards — BOM rule violations are now shown as expandable cards per rule, with affected item counts and suggested fixes — matching the pattern used elsewhere in Redline.

Sprint 19 — MVP test round 2 fixes

EC numbering — Engineering Changes now use their own number sequence (EC-YYYY-NNNN), separate from ECR and ECO numbers.

Responsible Engineer — New optional field on Change Requests (both create and detail). Dropdown filters by Engineering department. EC list resolves engineer names instead of showing IDs.

EC detail restructure — “BoM Upload” tab renamed to “Engineering Change Details”. Attachments moved from a separate tab into the bottom of this tab. Status stepper added below the breadcrumb.

BOM comparison — Source (item type) and Effective From columns added. Delete buttons removed (comparison is now read-only). CSV download includes revision and effective date columns. QTY column mapping fixed for snapshot imports.

CR governance — Return to Draft button on submitted ECRs. Timeline deduplication for approve/reject events. Info tooltips on Classification and Severity column headers.

Action lists — Assignee dropdown grouped by department (department users first, then Others). Action templates page has a department filter. Priority and department updates no longer flash/reset.

Sprint 18 — MVP polish & UI consistency

Access control & governance — Coordinator-only checks on ECR status transitions (Start Review, Submit Decision). CCB process can now be toggled on/off in Settings, hiding the CCB Review step from the workflow when disabled.

Filter system overhaul — All list pages (Change Requests, Engineering Changes, ECOs, ERP Items) now use multi-select filter dropdowns with checkboxes, Select All, Clear Selection, and badge pills matching the table. Archived records toggle on CR, EC, and ECO pages.

Preliminary action improvements — Priority column added to action lists with inline-editable badge dropdown. Department badges are clickable to change directly in the table. Add Action and AI Suggest buttons side by side. Due date column on EC/ECO action lists.

Dynamic department colours — Department colours are now stored in tenant settings and applied consistently across the entire application. New departments auto-receive a colour from a 12-colour palette. Departments in use cannot be deleted (usage check shows what is attached).

Table & pagination — Tables with 50+ rows show a pagination bar fixed at the bottom of the viewport with sticky headers. Default sort indicator shown on page load.

ECR detail polish — Title displayed next to ECR number, horizontal pill badge layout for classification/severity/trigger/change type, collapsible attachments section, Safety badge as red pill, info tooltips on Classification and Severity fields.

Sprint 17 — Phase-in calculator integration

Phase-in at the EC level — engineers can now run phase-in calculations directly on an Engineering Change, before an ECO exists. Select a strategy, click Suggest Phase-in Dates, and review per-item suggestions with conflict flags and action previews. Accept individually or all at once.

Phase-in suggestion engine on the EC Affected Items tab — strategy dropdown (RunDown, NewOnly, Stop, Proactive), suggest button, accept/reject per item
Action preview per item — rule-based action drafts shown inline (read-only preview, final actions generated on ECO release)
Enriched affected items table with real-time ERP data (stock, open POs, production orders, lead times)
EC approval workflow with approve, decline, and resubmit cycle
CSV column mapping popup on BOM upload — auto-matches headers, shows sample data for mismatched columns
Auto-delta calculation after BOM upload (no manual trigger needed)

UI improvements across all pages:

Search bars on every list page (Changes, Engineering, ECOs, Audit, all five ERP data pages) with info tooltip showing searchable columns
CSV export on all ERP data pages (Items, Stock, POs, Production Orders, Supplier Items)
Item type filter dropdown on ERP Items page (MAKE, BUY, ASSEMBLY, PHANTOM, DUMMY)
Theme toggle — three modes: dark, light, auto (system default). Collapsed sidebar shows a single icon for the opposing theme
Standardised badge system: status badges (dot + uppercase) vs value badges (no dot, capitalised) across all pages
Audit page: user IDs resolved to display names, event type badges in filter dropdown

Sprint 16 — BOM comparison & validation

Side-by-side BOM comparison view (ERP vs new BOM) with delta highlighting
Three structural validation rules (BOM_RULE_01, 02, 03) with inline fixes
BOM import state machine: IMPORTED → ERRORS_FOUND/CLEAN → STORED
Auto-push to ERP when validation passes (bom_auto_push_on_clean setting)
Violations displayed in separate containers per rule with re-validate action

Sprint 15 — Action list generation engine

Rule engine for seven action types (UPDATE_BOM, CANCEL_PO, AMEND_PO, UPDATE_PROD_ORDER_BOM, HOLD_PROD_ORDER, MBOM_REVIEW, SERVICE_ACTION)
AI-generated instruction text per action item via Claude API
Priority auto-assignment based on days to due date
Department-grouped accordion view with completion counters
Fallback templates when Claude API is unavailable
Regenerate action list from fresh ERP data

Sprint 14 — Notification system

N1 release notifications — one email per department assignee with grouped actions
N2 automated reminders after configurable delay (default 48 hours)
N3 escalation to L2 and L3 with configurable day thresholds
One-click email response — token-based status update, no login required
HMAC-SHA256 signed tokens with 72-hour expiry, single-use enforcement
Branded HTML email templates (dark theme, mobile-optimised)

Sprint 13 — Tasks & action item management

Tasks page for department users with open action items
All Tasks page for coordinators with search, filters, and sorting
Sidebar badge with open task count
Action item status workflow: Open → Acknowledged → In Progress → Done / Skipped / Blocked
Contact-based delivery for non-user email addresses

Sprint 12 — Dashboard & analytics

Dashboard KPI cards: active ECOs, completion percentage, overdue actions, upcoming phase-ins
Active ECOs table with strategy badges and action completion progress
Overdue actions table with days-overdue counter
Upcoming phase-ins table with conflict flag indicators
Change request task queue for coordinators

Sprint 11 — Data quality monitoring

Seven data quality checks (DQ-01 through DQ-07): missing lead time, staleness, orphaned items, duplicate suppliers, inconsistent MOQ, missing UoM, negative stock
Automatic detection on ERP sync
Dashboard display of active flags per tenant
Historical tracking with resolution timestamps
AI-enhanced pattern detection

Sprint 10 — ERP data hub

ERP Data Hub central dashboard with upload cards and BOM snapshots table
Item Master page — paginated searchable list with item details
Item Detail page — full history, stock by location, open POs, production orders, supplier items, where-used analysis
Supplier Items, Stock Levels, Open POs, Production Orders pages
Per-row error reporting on upload with suggested fixes

Sprint 09 — ERP connector abstraction

Abstract connector interface for future integrations
Dynamics 365 Business Central connector (API scaffold)
Exact Online connector (API scaffold)
Configurable sync intervals and staleness tracking
Connector configuration in Settings → ERP

Sprint 08 — ECO release & close

ECO release with phase-in date validation gate
Action list auto-generation on release
ECO close with BOM effectivity updates (effective_from / effective_to dates)
Auto-archive of parent EC on ECO closure
Released/verified/closed timestamps and user tracking

Sprint 07 — Phase-in calculation engine

Four deterministic calculation strategies: Run Down, New Only, Stop, Proactive
Lead time resolution with three priority modes (item master first, supplier preferred, supplier shortest)
Six conflict flag types: MISSING_LEAD_TIME, PO_AFTER_PHASE_IN, PROD_ORDER_AFTER_PHASE_IN, CONCURRENT_EC, FCO_REQUIRED, IMMEDIATE_ACTION
AI-generated reasoning explanation per item
Phase-in schedule persistence with manual override and reason tracking
Phase-out date auto-set for removed items

Sprint 06 — ECR affected items & attachments

ECR affected items with item ref, BOM level, source, quantity, UoM, and notes
Auto-population of child items from ERP BOM when top-level assembly added
File attachments — PDF, Word, Excel, PNG, JPEG, TXT (max 25 MB, up to 10 per ECR)
Editable display names on attachments
Signed URL generation with 1-hour expiry
AI-powered duplicate detection for ECRs (semantic similarity on title + description + affected items)
Preliminary action items with templates, manual entry, and AI suggestions
Carryover of affected items and preliminary actions from ECR to EC on approval

Sprint 05 — BOM & ERP data layer

BOM ingestion from CSV — full parent-child structure with where-used analysis
Lead time resolver — reads lead time from item master and supplier tables
BOM delta calculator — compares two BOM snapshots and surfaces affected items
Multi-mode BOM structure support: level_parent and unique_key
ERP data staleness enforcement — calculations blocked if data exceeds the configured threshold
Tenant isolation confirmed on all new BOM tables

Sprint 04 — ECO workflow & phase-in engine

Full ECO lifecycle: Draft → Released → In Progress → Verified → Closed
Phase-in calculation engine with all four strategies (Run-down, New-only, Stop, Proactive)
Strategy comparison view — side-by-side dates, conflicts, and inventory impact
Per-item phase-in with governing date recommendation across multi-item ECOs
AI-generated action list (Claude API) with per-line ERP reference and justification sentence
N1 / N2 / N3 notification system via email (Resend)
One-click email response — action status update without login required
Tenant settings: numbering formats, buffer days, staleness threshold, department configuration
Full audit trail on all ECO events — append-only, actor + timestamp on every entry

Sprint 03 — ECR, CCB & Auth

ECR intake form with configurable numbering scheme (e.g. ECR-{YYYY}-{NNNN})
Classification workflow: Minor / Major / Emergency with fast-track for minor changes
CCB decision recording with approval conditions and deferred handling
Supabase Auth with role-based access (Coordinator / Viewer / Department User)
JWT-based tenant scoping — tenant_id derived from authenticated session, never from client
Multi-tenant data isolation with Row Level Security (RLS) enforced at database level

Sprint 02 — ERP data ingestion

CSV ingester for all five canonical data objects (items, stock, open POs, open production orders, BOM)
Column name tolerance — variant column headers accepted and mapped automatically
Validation on upload with per-row error reporting
Upload snapshot labelling with timestamp
Tenant-scoped storage paths for all uploaded files

Sprint 01 — Foundation

FastAPI backend with service layer architecture (business logic never in routes)
Next.js 14 frontend with App Router and shadcn/ui
Supabase PostgreSQL with tenant_id on every table from day one
Railway (backend) and Vercel (frontend) deployment pipeline
GitHub repository and branch strategy established