Achilles
Book a call
Get a free stockout reportRefer a plantLog in

Forward-deployed engagements

Case studies.

Paid pilots and deployments where forward-deployed engineers turned the rules a plant runs on into software: BOMs from drawings, invoices matched to jobs, and live ERP-to-MES data.

Different plants, same problem: five systems that don’t talk, and a few people who know how to reconcile them.

Automated production line in a manufacturing hall
DwgACH-CS-01
ScaleNTS

Costed BOMs in hours, not weeks, on a custom lighting line A configurable-lighting manufacturer where every order is a unique build. Constructors spent two to three weeks rebuilding each order's BOM by hand, even though most components repeat.

Read the case
[01/04]Architectural & Custom Lighting3-week paid pilot

Costed BOMs in hours, not weeks, on a custom lighting line

Paid pilot · design partner

A configurable-lighting manufacturer where every order is a unique build. Constructors spent two to three weeks rebuilding each order's BOM by hand, even though most components repeat. Achilles proposes the BOM from the order itself, so constructors review and approve it in hours and parts ordering starts the same day.

Weeks → hours
BOM turnaround
80–90%
of lines auto-proposed
Same day
parts ordering starts
1
approver with the final say
The challenge

Orders arrive from a configurator as drawings and specs, but nothing connects them to production. Constructors rebuilt each element in CAD and typed the BOM into Excel by hand. Most of every BOM is standard components; only a few positions change from order to order. Until the BOM was finished there was no parts list to buy from, so purchasing waited weeks.

Our approach
  1. 01Read each order's configured spec, drawing and constraints sheet as it lands
  2. 02Mine rules from past approved BOMs to auto-propose the standard 80 to 90%
  3. 03Vision model resolves the few variable positions and flags any it can't confirm
  4. 04Constructor reviews only the flagged lines and approves; the human holds final authority
  5. 05Export the approved BOM to the Excel format purchasing already uses
Client

Architectural and custom lighting manufacturer with a configurable product line. Each order is a set of elements, and every element carries its own BOM.

Systems
Configurator export (PDF + DXF)Constraints sheetLibrary of approved BOMs
Architecture
DXF parsingRules mined from historyBedrock vision modelCelery + S3Tenant-isolated
Outcomes
  • BOM turnaround from two to three weeks to hours
  • Constructors redirect to the custom 10% that needs judgment
  • Parts ordering starts the day the order lands
  • Paid from day one: proof of revenue, not a free demo
[02/04]Discrete / Converter Manufacturing3-week fixed-fee PoC

From a 15-email chain to one approver per invoice

Paid proof of concept · design partner

A metal fabrication group where accepting a single invoice took a chain of about 15 emails and up to 10 people. Achilles reads each e-invoice, matches it to its order and job, and proposes the project, so only the ambiguous ones reach a single approver and the rest can clear without an email.

10 → 1
people per invoice (goal)
~15 emails
replaced by one queue
≥60%
auto-verify target
4
attribution patterns learned
The challenge

KSeF guarantees a tax-validated invoice but nothing about context: no project, no cost center, no order reference. Clearing one invoice today means logging into KSeF, emailing to ask if it matches the order, checking internal records, and only then accepting and assigning a project. About 15 emails and up to 10 people CC'd to clear a single invoice, hundreds of times a month.

Our approach
  1. 01Read KSeF invoice XML directly, no OCR: vendor, line items, amounts, and project tokens
  2. 02Match each invoice to order records and produce an order-match verdict
  3. 03Propose project and cost center with a confidence score using four learned patterns
  4. 04Seed the project dictionary from the client's own historical labels
  5. 05Route only low-confidence or mismatched invoices to one named approver, replacing the broadcast CC
Client

Metal fabrication group. Mandatory national e-invoicing via KSeF.

Systems
KSeF invoice XMLOrder recordsERP (system of record)Historical reply-email labels
Architecture
XML parser (no OCR)Attribution engine + confidenceReviewer queue + audit trailSynthetic-first; data stays on-site
Outcomes
  • Goal: from 10 people to 1 approver, on under 20% of invoices
  • Order mismatches caught before payment, not after
  • Clean, project-tagged cost data from day one
  • Phase 2 path: true per-project material cost and loss
[03/04]Industrial ManufacturingOne quarter to full production

From day-old reports to live plant data in under 30 minutes

Before Achilles · team track record

Took a $1B solar manufacturer from 24-hour reports to plant data under 30 minutes old and gave back 40 hours a week of manual reconciliation, so planners schedule against live stock and WIP.

<30 min
Data latency
99.5%
Pipeline reliability
40 hrs/wk
Time saved
<1%
Error rate
The challenge

Production metrics locked in Oracle MES behind VPN. Financial data scattered across Oracle ERP and SAP. 40+ hours weekly spent on manual reconciliation. 24-hour reporting delays blocking AI initiatives.

Our approach
  1. 01Automated VPN access and catalogued 328 Oracle schemas with latency mapping
  2. 02Activated change-data-capture (CDC) across Oracle and MES systems
  3. 03Built unified event schema preserving order-to-shipment causality
  4. 04Containerized self-healing pipeline with automated health monitoring
  5. 05Deployed real-time dashboards with LLM-powered translation for supplier docs
Client

U.S. solar technology manufacturer ($1B revenue).

Systems
Oracle 19c ERPSAP modulesLegacy MES via FortiClient VPNMandarin supplier PDFs
Architecture
Oracle LogMiner CDCAzure PostgreSQL event storeDocker containerized pipelinesGitHub Actions CI/CD
Outcomes
  • Finance reviews live dashboards instead of day-old spreadsheets
  • Planners schedule against real-time WIP visibility
  • Foundation enables demand forecasting and predictive maintenance pilots
  • Zero manual reconciliation required
[04/04]Energy & Battery Storage16 weeks to production

Order status in minutes, not three days, for grid-scale storage

Before Achilles · team track record

Connected a battery plant's ERP and shop floor so production and inventory were visible in real time: customer order updates went from 72+ hours to under 5 minutes, and order-to-delivery fell 35%.

Real-time
Order visibility
-35%
Cycle time
100%
Quality traceability
<5 min
Customer updates
The challenge

ERP system (SAP) couldn't communicate with shop floor MES. Production planners relied on end-of-day reports. Quality data trapped in isolated databases. Customers waiting 72+ hours for order status updates.

Our approach
  1. 01Built bidirectional integration layer between SAP ERP and proprietary MES
  2. 02Implemented real-time production event streaming to unified data warehouse
  3. 03Created cell-level quality traceability linking test data to production batches
  4. 04Deployed customer-facing portal with live order tracking and delivery ETAs
  5. 05Automated compliance reporting for grid-interconnection requirements
Client

Grid-scale battery manufacturer ($500M revenue) supplying utility-scale energy storage projects.

Systems
SAP S/4HANA ERPProprietary MESCell testing databasesSCADA systems
Architecture
Kafka event streamingSnowflake data warehouseReal-time API layerAutomated compliance reports
Outcomes
  • Production teams see live WIP across all assembly lines
  • Quality issues traced to specific cell batches within minutes
  • Customers self-serve order status instead of calling support
  • Compliance reports generated automatically for grid certifications

Start with one workflow

See what this looks like on your plant.

Every engagement starts with one broken workflow. We’ll map how it runs today and scope a paid pilot.

Book a callGet a free stockout reportSee the live demo
Case Studies · Achilles