One planning system for yarn, looms and the process house. It sits on top of SAP and gives every order a date it can keep.
Marketing enters an order in SAP. PPC360 checks yarn, looms and the process house together and gives a date it can keep. When the floor slips, it simulates options and a planner picks one.
PPC360 reads SAP and writes back only the plans a planner has published.
Each date shows the yarn, loom and machine slots behind it, in plain words.
PPC360 suggests and simulates. A named planner publishes, and can always override.
Today a planner makes each one by hand in Excel. SAP only records what was done.
The engine is normal code, so the same input always gives the same plan. AI is used only to explain, to answer questions and to score risk.
| What | How |
|---|---|
| Read Orders, BOM, stock, POs, confirmations, QA results | SAP’s standard interfaces (OData services, BAPIs or RFCs), read every few minutes. New and changed orders are picked up on their own. |
| Write back Dates, sequences, process orders | Through standard SAP interfaces, and only after a planner publishes a plan. |
| Middleware | D’Decor’s existing middleware if there is one. If not, a small connector service in D’Decor’s cloud. |
| Test first | Built and tested against a SAP test system before it touches production. |
We agree the exact method with D’Decor’s SAP team in discovery.
| Data | What we check | How we fix it |
|---|---|---|
| Material BOM | Yarn per metre matches what the floor actually issues | Compare with 3 months of SAP goods issues |
| Efficiency standards | 1 standard per product and loom type | Fill gaps from loom production history |
| Loom master | Type, width and what each loom can weave | Confirm with the weaving heads |
| Machine master | Minimum batch and capacity per machine | Confirm with the process house heads |
| Changeover times | Hours for knotting, gaiting and shade changes | Time them on the floor in discovery if they are not recorded |
| QA history | Rejection reasons coded the same way | Map free-text reasons to a short code list |
Each target comes from the brief. PPC360 tracks every one of them from the first week it goes live.
| Measure | Target in the brief | How PPC360 tracks it |
|---|---|---|
| OTIF delivery | At least 95%, aiming for 100% | Every order against the date confirmed at order entry |
| Date reliability | Every date from real capacity | Share of orders whose date came from PPC360, not a standard band |
| Loom efficiency | Fewer changeovers, more knotting, less gaiting | Gaiting and knotting count per month, and loom days lost to them |
| Process house use | Optimum machine use, with MOQ and clubbing respected | Load per machine and batches run below the machine minimum |
| Rejection and rework | Early alerts, fewer reprocess loops | Risk alerts raised, and metres reprocessed per month |
| Visibility | Live status for every stakeholder | Ageing by work centre, live order status and the weekly report |
1 week on site with the PPC team. Then order promising and the loom plan for 1 loom group, run beside Excel on D’Decor’s own orders.
Order promising, yarn requirement and the yarn lines, and the full month loom plan across every loom type.
Dyeing, printing and finishing plans with clubbing and minimums, alerts at every stage, and replanning with simulation.
Stitching for bedsheet orders, live status and the weekly report, and quality history that feeds the next order.
Planners compare both plans for 2 weeks, while the next phase is built, before Excel retires for that module.
Write-back to SAP starts module by module, once the planners sign off.
13 weeks from kickoff to the last module going live. We confirm the plan at the end of discovery.
| Risk | What we do about it |
|---|---|
| Gaps in master data Efficiency standards, BOMs, machine minimums | We check the data in week 1 and list every gap with an owner before any plan goes live. |
| Planners do not trust the plan | Every date and every move shows its reason. Planners can override with a note, and Excel runs beside PPC360 until they sign off. |
| Writing to SAP | We start read-only. Write-back uses interfaces agreed with D’Decor’s SAP team and starts only after sign-off. |
| Shop floor adoption | Each work centre sees only its own queue, on a floor screen or a phone, in the language its team prefers. |
| Rules we have not seen yet | Rules are settings in the engine, so a new finish, loom type or machine is added without a rebuild. |
The brief leaves these open. Each one changes how the planning engine is set up.
SAP version and modules, and whether plans post as planned orders, production orders or a custom table.
Looms by type and width, and which constructions need knotting or gaiting.
Are standard efficiencies held per product in SAP, or only in Excel?
Hours for knotting, gaiting and shade changes on each machine.
Who clubs greige, and how far ahead of the delivery date.
Machine list, MOQ for each, and the finish order rules in use.
Share of lots reprocessed today, and how rework is booked in SAP.
How OTIF is measured now, with figures for the last 6 to 12 months.
Order promising and the loom plan for 1 loom group, on D’Decor’s own orders.
Modules 1, 2 and 3, with SAP read and write-back for them.
Modules 4, 6 and 7: process house, alerts, replanning and simulation.
Modules 5 and 8: stitching, live status, the weekly report and quality feedback.
Excluding GST. The final figure is agreed before go-live.
D’Decor and 360 Labs
SAP extracts, Excel planning files, loom and machine master, BOM samples, standard efficiencies, sample orders and rejection history.
360 Labs
Time with the PPC team and work centre heads, and read access to SAP in a test system.

BSc Mathematics from the University of Toronto, with modules in computer science, psychology, philosophy and advanced economics, followed by BSc IT and Digital Innovation at City University of London, specialising in AI, machine learning and digital product design.
Lead AI Product Manager and Head of Growth at brdge.ai in London, scaling the company past $1.5M ARR in under 6 months alongside the core founding team and delivering AI projects for Barclays, Leonardo, the NHS and the UK Home Office. Core member of the product and growth team at DeepReel.com in London, valued at $10M+, through its $1M pre-seed round, leading enterprise sales and delivery for Accenture and the Aston Martin Aramco F1 Team. Co-founded and raised funding for 2 AI SaaS startups before founding 360 Labs.

Hardware engineer and systems programmer with 12 years building government-scale and consumer-scale systems. Built real-time instrumentation for the Indian Meteorological Department and engineered core systems behind IDFC Bank's customer banking application, serving 1.2 million users. Contributed the browser module to BOSS Linux, India's national operating system, deployed across 6M+ government, education and defence devices for C-DAC. Architected SLM360, India's first sub-50MB language model, running on-device at 39ms latency.
Indian patent holder, published author, with Med360 and AgentGuard released publicly. Quantum in Machine Learning and AI Systems, IIT Delhi.

Lawyer and legal technologist. Qualified as Solicitor (England and Wales) from Herbert Smith Freehills Kramer (London), advising UBS, Credit Suisse and Morgan Stanley on roughly $10Bn of transactions and disputes. Founded the firm's Digital Committee, shipping AI tooling into live matters including various $1Bn+ matters. Involved in procurement and pilots on Harvey, Legora, Orbital and other AI vendors. LLB, LLM, MSc AI from various UK universities.

Senior Software Associate at Capital One, where he led backend engineering for a platform launched across North America at NADA '26. Received multiple Director awards and Team awards. B.Tech and M.S. by Research in Quantum and AI from IIIT Hyderabad, with a first-authored publication in the Journal of Physics A, a patent in progress, and AWS Certified Solutions Architect.

Technical Product & Project Leader building 0→1 SaaS platforms across fintech, CRM, ERP and cybersecurity. Former Sole Product Lead at MortgagX, an AI-native UK mortgage platform engaging Lloyds and Nationwide Building Society. Delivered 200+ production releases with 0 critical defects and 300+ third-party integrations. Led a cybersecurity programme for Belgian firm Sertalink. At Van Technologies, managed UK and European operations and built an early online load-planning system for the removals industry.

Former Senior Engineering Leader with 5+ years at Red Hat, leading Fedora Linux Release Engineering for 1.2M+ users, managing 50,000+ daily packages across 8 releases. Founding member of Konflux, powering OpenShift. Recipient of 2 Red Hat Impact Awards, with 10+ years contributing to Debian and mentoring GSoC & Outreachy. Published IEEE researcher and speaker at FOSDEM, Flock, DevConf, GNOME Foundation and FOSSASIA. CNCF-certified, Product Owner certified by Scrum Alliance, with experience at Nokia and an MBA from NMIMS.

Full-Stack AI Engineer with 5+ years deploying machine-learning systems into production, across real-time systems and full-stack LLM applications. Former Senior Software Engineer at Alegance, where he built AI web applications and production LLM integration pipelines. His experience spans retrieval-backed systems, AWS infrastructure and secure enterprise platforms. Open-source contributor to Neovim (80K+ GitHub stars), Palantir's Blueprint, Hermes Agent, Alacritty, OpenAI, Unsloth, ngrok and Open Jarvis, with a focus on developer tooling and AI agent frameworks.

Full-Stack AI Engineer specialising in data ingestion, retrieval and multimodal AI: turning raw, inconsistent enterprise data into records that models and planners can trust. Selected for Amazon ML Summer School, Amazon's selective machine-learning programme in India, and for McKinsey.org Forward, its programme for early-career professionals. Placed 2nd nationally at the Samsung PRISM Hackathon, run by Samsung R&D Institute India. Contributes to Gemini open source and Google Summer of Code. His focus is AI that reads documents, images and structured data together.
We have engineers on 4 continents. We build for Nifty-500 and Fortune 500 companies, and our logistics platforms are live in India and the USA.

This presentation is built for a wide screen. Rotate your phone to landscape to view it.
If the screen does not turn, switch on auto-rotate