Chat with us
India · Malaysia · USA — 65+ engineers, incorporated february 2024

Intelligent enterprise engineering.

We build and run the enterprise applications, applied AI, and intelligent automation that mission-critical operations depend on — for clients across FinTech, healthcare, supply chain, freight, and IT services, from delivery centres in India and Malaysia.

00 / Thesis

PMDG engineers intelligent systems that take work off people — and runs them in production.

In production now Delivered results
2,800+
Enquiries handled per day
Global & regional freight forwarder (Canada & India)
75%
Cost saving on physical verification
Pharma supply chain company (India & Middle East)
100%
Offline continuity for field operations
National electricity utility — Malaysia
See all fourteen engagements
By the numbers
65+
Engineers and delivery staff
25+
Enterprise clients
25+
Projects delivered
5
Offices, India and Malaysia
3
Countries with entities

India, Malaysia, United States

Trusted by Teams at Over 100+ Enterprise Manufacturing & Industrial Operations
  • Asian Football Confederation
  • Mitsubishi Heavy Industries
  • Choithrams
  • MegaMart
  • RMedisys
  • My Dream Gold
  • Sharjah Ladies Club
  • Flow Supply Chain
  • Ceres India
  • EIL Global
  • CloudIntuit
  • WonderWrks
  • Spartan Equipment
  • aiClaim
  • Tenaga Nasional Berhad
  • PharmaCare Logistics
  • Accounts NextGen
  • Scorpion
  • Avaan
  • Royal Tech Systems
Enterprise & Industrial ERP Connectivity:
SAP (S/4HANA & ECC) Oracle NetSuite Microsoft Dynamics 365 Siemens & Rockwell Infor / Epicor
01 / Capability What we build

8 capabilities, each removing a specific cost.

Every one exists because a client had a process that was expensive, slow, or unreliable. They are listed by what they take away, not by the technology behind them — pick one for the detail.

Automation — RPA, BPA & Agentic AI

Take the repetitive work off people, and keep it off.

We map the process as it actually runs — not as the SOP describes it — then automate the parts that never needed a human. Bots handle the clicks, document AI handles the reading, and agents handle the judgement calls that can be safely delegated. Six dedicated RPA engineers, and automations live in production processing thousands of transactions a day.

What you get
Process discovery and time-and-motion baseline
Attended and unattended bot development
Document capture, OCR/ICR and LLM extraction pipelines
Agentic workflows with human-in-the-loop review queues
Run-state monitoring and bot health alerting
Delivered with
UiPath Automation Anywhere Python DoIT SAP Business One

Enterprise App Development

The system of record your operation actually runs on.

Secure, scalable line-of-business applications built with agile delivery — role-based access, audit trails, integration with what you already run, and a data model that survives the business changing shape. From omnichannel commerce across five countries to warehouse management across multiple sites.

What you get
Requirements and process modelling
Custom platform build, single codebase across regions
Migration from legacy and spreadsheet-based systems
Role-based access control and audit logging
Multi-currency, multi-language and multi-entity support
Delivered with
Java/J2EE React .NET Node PHP PostgreSQL

ERP Implementation

Oracle and SAP configured for how you actually work.

ERP programmes fail when the package is implemented as shipped and the business is asked to adapt. We configure to business fit, integrate into the systems around it, and automate the processes the ERP was never going to cover on its own — order management, three-way matching, reconciliation, month-end close.

What you get
Fit-gap analysis and configuration to business process
Oracle, SAP and SAP Business One implementation
Data migration and cutover planning
Process automation layered on top of the ERP
Post-go-live hypercare and support
Delivered with
SAP SAP Business One Oracle Zoho Nexus360

Data Engineering & Science

Governed data that turns into a decision, not a slide.

Most analytics work stops at the dashboard. We build the step after it: a single source of truth across fragmented ERPs, a common reporting schema that does not require ripping out the source systems, automated master-data management to resolve duplicates, and models whose output triggers something and is measured against what actually happened.

What you get
Cross-entity data integration and common GL standard
Automated MDM and data-quality firewall
Pipelines, warehousing and real-time streaming
MLOps: deployment, versioning, drift monitoring, retraining
Operational and financial dashboards
Delivered with
Spark Kafka Airflow MLflow CDC Power BI Tableau

Cloud Engineering & Support

Infrastructure someone is accountable for at 3am.

Migration, landing zones, CI/CD, and the observability to know something broke before a customer tells you. Where availability is the requirement rather than a preference, we design for it explicitly — including offline-first field operations and high-availability disaster recovery.

What you get
Cloud migration assessment and landing zone design
AWS, Azure and GCP delivery and optimisation
Containerisation and orchestration
HADR architecture and failover testing
Monitoring, alerting and on-call runbooks
Delivered with
AWS Azure GCP Docker Kubernetes

Enterprise Integration

Your systems stop disagreeing with each other.

Middleware, API gateways and sync layers that connect line-of-business applications into core systems and ERPs — including the ones with no usable API. Built to decouple, so a field application does not go down because the core system is being patched.

What you get
Integration architecture and API gateway design
Microservices middleware and message-based sync
Two-way, near-real-time and scheduled synchronisation
Single sign-on and federated identity
Full audit trail and traceability across systems
Delivered with
Spring Boot REST/SOAP Kafka IBM Maximo ServiceNow JIRA

Mobile Development

Software for the people who are not at a desk.

Field engineers, warehouse operators, drivers and clinicians need applications that work on a poor connection, on a mid-range Android, in one hand. We build offline-first and treat sync conflicts as a design problem rather than an edge case — one utility field app runs at 100% offline continuity.

What you get
Cross-platform and native application development
Offline-first data sync and conflict resolution
Device integration: camera, barcode, GPS, signature, biometrics
Store submission, phased rollout and crash triage
Ongoing maintenance and OS-version upgrades
Delivered with
React Native Swift Kotlin Firebase

Quality & Support

Human-verified assurance, not a green pipeline badge.

A dedicated QA and support function — seven people — covering functional, integration and regression testing, plus managed application support once the build is done. Automations we deploy ship with health checks, because an automation that fails silently is worse than the manual process it replaced.

What you get
Functional, integration and regression test suites
Automated pipeline gates and release verification
Managed application support and SLA reporting
Incident triage, root-cause analysis and runbooks
Continuous improvement backlog
Delivered with
Selenium Playwright JIRA ServiceNow
02 / The loop How it all connects

Every engagement is the same seven moves.

Different sector, different systems, same shape. This is the loop we build, and the reason an automation programme with us starts from a working pipeline rather than a blank page.

01

Signal arrives

An order by EDI, an enquiry by email, a form submission, a scanned document, a reading from a machine. Work enters your business in whatever shape the sender chose — rarely the shape your systems want.

02

It gets understood

OCR, ICR and language models turn the unstructured thing into structured fields. This is the step that used to be a person reading one screen and typing into another.

03

A decision is made

Rules check it against your own policy — credit limits, entitlements, thresholds, contractual terms. A model scores what rules cannot express. Anything genuinely ambiguous routes to a person; the rest does not.

04

Something happens

A bot creates the record, posts the entry, raises the work order, sends the response. Where an API exists we use it; where one does not, RPA does the clicking.

05

The system of record updates

SAP, Oracle, Dynamics, Maximo, Nexus360 — whatever you run. One write instead of four re-keyings, with the audit trail attached to the action rather than reconstructed afterwards.

06

It gets measured

Cycle time, exception rate, straight-through rate. Not a dashboard for its own sake — the numbers that tell you whether the automation is still earning its place.

07

And it feeds back

Rejections become new rules. Exceptions become new training data. The loop tightens every week instead of ossifying the day we hand over.

01 / Signal arrives 1 of 7
03 / Demonstration The transformation canvas

Pick your sector. Watch where the time goes.

Rather than describe automation, this draws it. Choose the operation closest to yours and the canvas builds the process as it runs today, then rebuilds it the way we would deliver it.

“Enquiries arrive faster than we can quote them”

How it runs today 1h 19m
Enquiry lands by email, WhatsApp or portal 3m
Read and interpret an unstructured request 9m
Extract shipment detail by hand 11m
Request rates from each carrier 18m
Wait for replies, compare them manually 24m
Assemble and send the quotation 14m
How it runs after 6m
Multichannel capture across every channel 12s
LLM and ICR extract the shipment detail 30s
Validate against lane and carrier rules 18s
RPA co-pilot collects rates across carriers 1m
Rates compared, quotation assembled 24s
Pricing desk reviews exceptions only 3m
Quotation delivered 6s
Time returned per case
92% 1h 19m → 6m
Delivered result 2,800+ enquiries handled daily · 70%+ faster invoice processing

Process shape and outcome are from delivered work; per-step timings are modelled, not measured. Your baseline is measured in discovery.

“Orders arrive as email and get typed into SAP”

How it runs today 52m
Purchase order arrives by email or EDI 2m
Read and interpret the order 8m
Check item codes and availability 12m
Verify customer and credit limit 9m
Key the order into SAP 15m
Confirm back to the customer 6m
How it runs after 4m
Email and EDI ingested automatically 12s
Order detail extracted 24s
Item, quantity, customer and credit validated 30s
Missing items and customers auto-created 36s
Sales order created in SAP Business One 18s
Ops clears flagged exceptions only 2m
Time returned per case
91% 52m → 4m
Delivered result 65% cost saving on order processing · near-zero manual errors

Process shape and outcome are from delivered work; per-step timings are modelled, not measured. Your baseline is measured in discovery.

“Reconciliation runs late and holds up the close”

How it runs today 1h 21m
Download the bank statement 4m
Read remittance instructions 13m
Pull incoming payments from SAP 8m
Match line by line 26m
Investigate the mismatches 19m
Correct the entries in SAP 11m
How it runs after 5m
Statement and remittance ingested 18s
Detail extracted and normalised 30s
SAP incoming payments retrieved 18s
Validate bank vs. remittance vs. SAP 36s
Incoming payment entries auto-corrected 24s
Finance reviews genuine breaks only 3m
Time returned per case
94% 1h 21m → 5m
Delivered result 55% cost saving · timely reconciliation and settlement

Process shape and outcome are from delivered work; per-step timings are modelled, not measured. Your baseline is measured in discovery.

“Tickets, approvals and invoices are three different tools”

How it runs today 1h 12m
Ticket raised in the service desk 3m
Triage and route to the right queue 11m
Chase the approver by email 21m
Re-key detail into the billing system 13m
Reconcile against the vendor invoice 17m
Close the ticket and update JIRA 7m
How it runs after 3m
Ticket captured and classified 18s
Routed by workflow rules 12s
Approval requested and tracked automatically 18s
Approver acts in one click 1m
Invoice matched and posted 30s
ServiceNow and JIRA updated in step 12s
Time returned per case
96% 1h 12m → 3m
Delivered result Ticket, approval and invoice flows unified on one workflow engine

Process shape and outcome are from delivered work; per-step timings are modelled, not measured. Your baseline is measured in discovery.

“Claims get denied and we find out weeks later”

How it runs today 1h
Retrieve encounter documentation 6m
Read notes, assign billing codes 14m
Check payer-specific rules by hand 9m
Key the claim into the billing system 7m
Submit in the overnight batch 2m
Denial returns, rework and resubmit 22m
How it runs after 4m
Encounter note ingested on close 12s
GenAI coding extracts billable detail 30s
Payer rule validation pre-submission 18s
Denial-risk model scores the claim 12s
Coder clears flagged claims only 3m
Clean claim submitted 6s
Time returned per case
93% 1h → 4m

Illustrative capability model, not a delivered result. Timings are typical for the process shape shown; your baseline is measured in discovery.

“Every claim is a pile of documents someone has to read”

How it runs today 1h 9m
Claim submission arrives with attachments 3m
Open and read each document 18m
Key claim detail into the portal 12m
Chase the claimant for what is missing 20m
Route for assessment and approval 9m
Update the policy record 7m
How it runs after 5m
Submission and attachments ingested 18s
Documents classified and extracted 36s
Completeness checked against policy rules 18s
Missing-document chase runs itself 12s
Assessor reviews the exceptions 4m
Policy record updated automatically 12s
Time returned per case
92% 1h 9m → 5m

Illustrative capability model, not a delivered result. Timings are typical for the process shape shown; your baseline is measured in discovery.

Put your volume through it Estimate
50 cases / day

Based on 250 working days and 1,650 productive hours per FTE per year.

55
Hours returned per day
13,750
Hours returned per year
8.3
Full-time equivalents freed
Effort removed Effort remaining
04 / Platforms Products & frameworks

We are not assembling every engagement from scratch.

Four platforms built and maintained in-house. They are why an automation programme starts from a working capture-understand-automate pipeline rather than from an empty repository.

AI-powered enterprise ERP suite

Nexus360

HRMS, finance, accounting, projects, payroll and support on one intelligent core.

A unified ERP suite for organisations whose back office has grown into four disconnected systems and a reconciliation spreadsheet. Built AI-first rather than retrofitted: ledger posting, three-way matching and month-end close are automated by default, not sold as an add-on module.

  • HRMS and payroll with statutory compliance
  • Finance, accounting and automated ledger posting
  • Project and resource management
  • Integrated support desk
  • Predictive analytics command centre
GenAI business process automation platform

DoIT

Capture, understand, automate, analyse, optimise — across Finance, HR, Ops and IT.

The platform underneath most of our automation work. It ingests documents and messages from any channel, understands them with OCR/ICR and LLM extraction, drives the downstream process, and reports on what it did — with compliance and audit built in rather than bolted on.

  • Multichannel capture: email, EDI, WhatsApp, web, portals
  • AI/LLM and ICR extraction with validation
  • Workflow and approval automation
  • Spend, cycle-time and exception analytics
  • Audit trail and compliance controls
AI assistants & autonomous agents

Agentic AI Frameworks

Workflow-aware, multilingual agents that act securely on your data.

Assistants and autonomous agents that do more than answer questions — they hold context about the workflow they sit inside, take actions against real systems, and operate under the same access controls as the humans they work alongside.

  • Workflow-aware assistants grounded in your systems
  • Autonomous agents with scoped, auditable permissions
  • Multilingual interaction
  • Human-in-the-loop escalation paths
  • Secure retrieval over enterprise data
Inventory & billing for small business

Bill360

Lightweight stock, invoicing and billing for small and mid-scale businesses.

Deliberately small. Not every business needs an ERP programme, and Bill360 exists for the ones that need stock and invoicing to simply work without a six-month implementation.

  • Stock and inventory tracking
  • Invoicing and billing
  • Customer and vendor records
  • Straightforward reporting
05 / Domain Industries

We are not generalists pretending to know your sector.

Six operations we have delivered into. Each one is listed with the specific failure we keep being called in to fix — if yours is not described here accurately, we are probably the wrong firm.

01 fintech

FinTech

Accounting, taxation and onboarding run through queues that each add a day, and booking-to-accounting is re-keyed by hand between systems that will not talk to each other.

Our approach

AI and RPA across onboarding, booking-to-accounting and reconciliation — with document verification and validation running in one pass, and only genuine exceptions reaching a reviewer.

02 healthcare

Healthcare

Revenue leaks between the clinical note and the paid claim. Coding is manual, denials arrive weeks later, and nobody can say which payer is rejecting what.

Our approach

Revenue cycle management with GenAI medical coding, conversational AI for patient interaction, and mobile applications for the clinicians who are not at a desk.

03 retail-supply-chain

Retail, Supply Chain & Manufacturing

Stock, pricing and product data live in systems that disagree, orders arrive as unstructured email, and returns are verified by eye at the receiving bay.

Our approach

Omnichannel platforms on a single codebase, RPA across order and returns processing, Industry 4.0 AI/ML and IoT for physical verification, and predictive analytics that reaches buying in time to matter.

04 logistics-freight

Logistics & Freight

Enquiries arrive across email, WhatsApp, web and carrier portals faster than anyone can quote them, and every hour of delay is business lost to whoever replied first.

Our approach

AI multichannel capture, LLM and ICR extraction from unstructured documents, automated rate collection and comparison across carriers, and quote and invoice automation.

05 itsm

IT Service Management

Tickets, approvals and invoices are three workflows in three tools, and the coordination between them is somebody's full-time job.

Our approach

ITSM platforms with automated ticket handling and approval routing, invoice automation, and integration across ServiceNow, JIRA and FieldCode.

06 insurance

Insurance

Claims workflows are document-heavy and deadline-driven, and the portal handling them is a decade-old application nobody wants to touch.

Our approach

Liferay application development and maintenance, claims workflow automation, and document extraction that turns unstructured submissions into structured claims data.

06 / Evidence Selected work

Live in production, with the numbers attached.

Fourteen engagements. Filter by sector, open any one for the full situation, approach and outcome. Figures are PMDG's own published results — each sourced, and anything unsourced withheld rather than estimated.

Showing 14 of 14 engagements

75%
Code reuse across Referee and Sports-Med modules
The situation

The federation's education and certification platform was a legacy PHP application carrying course, user, role, scoring and certificate management for referees and sports medicine programmes across the continent. It could not scale, and each new module was being rebuilt from scratch.

What we did

We re-engineered it onto a modern, cloud-native microservices stack on AWS behind a unified API gateway with single sign-on. The education management model was rebuilt as a standard reusable framework, with role-based access, encryption, data masking and audit logging throughout, and PDPA/GDPR-compliant data governance designed in rather than retrofitted. We added an AI-driven media layer that recognises and tags players, events and trophies automatically.

What changed

One reusable framework now serves multiple programme modules instead of one codebase per module, the platform is compliant by design, and media search and delivery are dramatically faster at scale.

AWS Microservices API Gateway SSO AI vision PHP migration
100%
Offline continuity for field operations
The situation

A national electricity utility needed IBM Maximo in the hands of field crews — but field work happens in places with no signal, and coupling a field application directly to the Maximo core meant outages upstream became outages at the pole.

What we did

An offline-first React Native application on top of IBM Maximo handling work orders, GPS and signature capture, notifications, and fault and supply-disruption reporting. Behind it, a Spring Boot microservices middleware layer decouples field operations from the Maximo core, with near-real-time and scheduled two-way sync and HADR on AWS. Licensing cost was reduced through multiplexing.

What changed

Crews keep working through any network condition, every field action carries a full audit trail, and the integration backbone now serves as the platform for future systems rather than a point solution.

React Native IBM Maximo Spring Boot AWS HADR
2800+
Enquiries handled per day
70%+
Faster invoice processing
100%
Extraction and validation accuracy (approx.)
The situation

Enquiries, pricing requests and shipment documents arrived across email, WhatsApp, web and carrier portals in volumes no team could keep up with. Rates had to be collected and compared by hand across carriers, quotes and invoices took heavy manual effort, and business was lost to whoever replied first.

What we did

AI multichannel enquiry capture feeding an LLM and ICR extraction engine, automated quotation and rate intelligence, and an RPA pricing co-pilot that collects and compares rates across MSC, Maersk and CMA CGM. Invoice automation with compliance validation closes the loop.

What changed

Enquiries are answered outside business hours, rate comparison is immediate rather than a research task, and revenue leakage from slow or missed quotes is substantially reduced.

AI/LLM extraction ICR RPA DoIT Carrier portal integration
75%
Cost saving on physical verification
The situation

Physical verification of medicines at receiving and returns was manual, error-prone and slow — and it buckled under peak-hour surges. Manual validation also carried real fraud and compliance exposure in a regulated supply chain.

What we did

An IoT conveyor with box cameras scans medicines as they move. AI performs item detection and verification, returns are compared automatically against the original claim request, credit-note approval is automated, and the whole flow integrates with the ERP.

What changed

Verification is automated and fraud-resistant, manual errors are near-eliminated, and the platform absorbs demand surges instead of queueing behind them.

IoT Computer vision AI detection ERP integration DoIT
65%
Cost saving on order processing
The situation

Order processing depended on people reading emails and EDI messages and keying the contents into SAP. Peak-hour surges caused backlogs, data-entry errors caused disputes, and processing delays turned into late deliveries.

What we did

End-to-end intelligent automation of sales order management in SAP Business One: read the email or EDI message, extract the order detail, validate item, quantity, customer and credit, then auto-create items, customers and orders in SAP. A real-time portal tracks order status throughout.

What changed

Order processing is elastic under surge, errors are near-eliminated, and the same extraction and validation pattern has been reused for invoice automation.

SAP Business One RPA EDI AI extraction DoIT
65%
Cost saving across the quotation process
The situation

Enquiry-reply delays were costing business outright. The quotation process depended on manual document handling and vendor chasing, and manual errors compounded the delay.

What we did

AI-powered enquiry automation across the quotation process — automated email and document processing, smart vendor interaction with partners, quote aggregation and optimisation, and automated response delivery.

What changed

Responses go out consistently and on time, turnaround is faster, and the workflow scales through demand surges rather than queueing.

AI document processing RPA DoIT Email automation
55%
Cost saving on reconciliation
The situation

Reconciliation ran late and created accounting friction downstream. Statements and remittance instructions were read by hand, matched by hand against SAP, and corrected by hand.

What we did

Automated bank reconciliation in SAP Business One: read bank statements and remittance instructions, retrieve SAP incoming-payment detail, validate bank against remittance against SAP, and auto-correct incoming payment entries.

What changed

Reconciliation and settlement happen on schedule, manual errors are near-eliminated, and peak volumes no longer extend the close.

SAP Business One RPA Document extraction DoIT
65%+
Reduction in manual admin effort
99.9%
Payroll and tax-filing accuracy
50%
Faster support response
The situation

HR, Finance, Accounting and Support had grown into data silos. Payroll and invoice processing were manual, repetitive work was driving high support latency, and there was no real-time financial visibility anywhere in the business.

What we did

A single intelligent ecosystem on the DoIT platform: intelligent HRMS and payroll with statutory compliance, AI-driven ledger posting and three-way matching, automated month-end close and tax reporting, AI transcription turning meetings into action items, and a predictive analytics command centre.

What changed

The back office runs as one system with real-time financial reporting, and the support function spends its time on exceptions rather than repetition.

DoIT Nexus360 AI/ML Payroll compliance Predictive analytics
100%
Audit compliance across the procurement workflow
The situation

Large-scale procurement ran on high-touch manual processes — fragmented tender and bid workflows, data-entry errors across RFPs and RFQs, slow multi-level approval cycles, and limited audit visibility.

What we did

A unified automated ecosystem on the DoIT platform: AI OCR/NLP extraction from proposals, weighted-matrix scoring and simulation, side-by-side bid battlecards, AI meeting intelligence that captures commitments, and ERP integration for purchase orders and three-way matching.

What changed

Procurement cycle times accelerated, AI risk scoring flags gaps before they become problems, and spend analysis and savings are visible on a dashboard rather than reconstructed after the fact.

DoIT OCR/NLP AI scoring ERP integration
3
Countries added on the same platform
The situation

A retailer operating across Dubai, Qatar, Kuwait, Bahrain and other UAE markets needed one omnichannel platform. Registration friction was suppressing acquisition, regional taxes and billing rules were managed through disconnected ERP logic, country-specific merchandising rules were hard to enforce, and there was no centralised customer insight.

What we did

A single-codebase, configurable omnichannel platform for multi-region use, with a low-friction registration flow offering multiple login methods, real-time ERP integration syncing products, inventory, pricing, taxes and customer data, and a centralised pricing engine handling regional taxes, discounts, loyalty and store-specific pricing.

What changed

The business expanded across markets on one platform instead of one build per country, merchandising rules are configuration rather than code, and customer intelligence now drives targeted marketing and retention.

React Node ERP integration Pricing engine Multi-currency
40%
Increase in operational efficiency
50%
Reduction in picking errors
The situation

High-volume, multi-location storage, picking, packing and distribution ran on manual, legacy-tool-heavy processes with limited real-time inventory visibility, data silos across sites, and little shipment traceability.

What we did

A web-based warehouse management system with real-time inventory tracking, barcode and mobile-enabled data capture, optimised picking, packing and shipping routines, and integration across ERP, mobile and third-party systems — built for multi-location scale from the start.

What changed

Inventory accuracy and workforce productivity both improved, order turnaround and dispatch accelerated, and operations run on analytics rather than intuition.

React Java/J2EE Barcode/mobile capture ERP integration
6 months
Foundation to go-live across four divisions
The situation

Reporting was fragmented and manual across four divisions running Dynamics 365, SAP and legacy MYOB. Every consolidated number had to be rebuilt by hand, and duplicates across entities meant no figure was fully trusted.

What we did

A centralised, governed, real-time analytics platform: cross-entity integration across D365, SAP and MYOB; a common GL standard giving one reporting schema without disrupting the source ERPs; and an automated master-data-management layer resolving duplicates before they reach a report. Delivered in three phases over six months.

What changed

One trusted reporting layer across all entities and geographies, with interactive dashboards turning consolidated data into real-time operational and financial insight.

Dynamics 365 SAP MYOB MDM Power BI Spark
The situation

Ticket handling, approval routing and invoice processing lived in three different tools, and keeping them in step was effectively somebody's full-time job.

What we did

A J2EE and React ITSM application with ticket management and workflow automation, automated approval routing and invoice automation — integrated across the client's existing ecosystem including ServiceNow, JIRA and FieldCode.

What changed

Tickets, approvals and invoices move through one workflow engine, and the coordination work between the three tools disappeared rather than being redistributed.

Java/J2EE React Workflow engine ServiceNow JIRA FieldCode
24/7
Continuous visitor engagement
The situation

A large hospitality facility needed personalised guest engagement and indoor navigation across multiple locations, with real-time visibility for management.

What we did

High-resolution cameras identify visitors and capture photographs automatically, composing a collage or reel of their visit delivered by email or WhatsApp. An interactive facility map provides point-to-point routing with live tracking, proximity push notifications and nearby points of interest, feeding a real-time management dashboard.

What changed

Guests get a personalised, navigable experience, management gets multi-location visibility, and the same vision platform extends to traffic monitoring and licence-plate recognition.

Computer vision Indoor positioning React Native WhatsApp API
07 / Terms What we commit to

Six promises you can hold us to in writing.

We would rather be judged on terms we have staked something on than on quotes you cannot verify. Every one of these is a clause we will put in the engagement letter.

01

We measure before we automate

Every engagement opens with a baseline of the process as it actually runs. If the measurement shows automation will not pay for itself, we say so and we stop. You keep the baseline either way.

02

You own the source, the infrastructure, and the runbook

Code lives in your repository, cloud resources are provisioned in your tenancy, and handover includes the runbook. Nothing we build requires PMDG to keep it running.

03

Fixed-scope discovery, before any build commitment

The first phase is a fixed-price, fixed-duration discovery that ends in a written technical plan and a costed build. You can take that plan to another firm. Some do.

04

One named engineer, accountable end to end

You get a technical lead who was in the discovery and is still there at handover — not an account manager relaying questions to a team you never meet.

05

Automations ship with monitoring, or they do not ship

Every bot and pipeline we deploy has health checks and alerting from day one. An automation that fails silently is worse than the manual process it replaced.

06

Accessibility and performance are acceptance criteria

WCAG 2.2 AA and Core Web Vitals thresholds are written into the definition of done, tested before sign-off. Not a phase two.

08 / Method How we work

Measure first. Commit second. Build third.

Most failed automation projects were scoped before anyone timed the process. Ours cannot be — discovery is a fixed-price phase that ends in a written plan you are free to take elsewhere.

3–6 weeks from first measurement to a proven pilot — everything before you are asked to fund a build.

  1. 01

    Measure

    We sit with the people doing the work and time it. The output is a baseline: volumes, cycle times, error rates, and where the cost actually sits.

    1–2 weeks
  2. 02

    Model

    A written technical plan — target architecture, integration points, the automation boundary, and a costed build with the assumptions stated.

    2–3 weeks
  3. 03

    Prove

    One process, end to end, in production with real users. Small enough to be reversible, real enough to prove the numbers.

    3–6 weeks
  4. 04

    Build

    The remaining scope, delivered in two-week increments against the plan, each one deployable.

    Engagement-dependent
  5. 05

    Hand over

    Runbooks, monitoring, access, and training for your team. Measured against the same baseline we started with.

    2 weeks
  6. 06

    Run

    Support and continuous improvement if you want it. Structured so you can end it without the system stopping.

    Optional, renewable
09 / Start Contact

Tell us what the process costs you today.

The useful first conversation is about a specific operation, not about our capabilities. Bring one process you think is too expensive and we will tell you honestly whether it is worth automating.

We reply within one working day, from a person, not a sequence.
Direct
Offices
  • IN India
    Trichy
    Corporate office · KK Nagar
  • IN India
    Trichy
    Development centre · EDA Street
  • IN India
    Coimbatore
    Development office · Fort
  • MY Malaysia
    Kuala Lumpur
    Regional office · Menara Shell, Brickfields