Capabilities

Engineering that runs the business, not just the website.

OneLab designs, builds and operates the systems enterprises depend on daily — AI-native operations, hardened infrastructure, resource planning, automation, bespoke software and the technology behind modern publishing.

50+
years collective experience
6
core capability practices
7
products built in-house
24/7
operate & support
01 — Capability

AI-powered business operations

Systems that decide, not just record.

We design, build and deploy AI-native operating systems that sit inside your workflow — scoring, routing, drafting and forecasting where the work actually happens.

Most AI projects stall because the model never touches the operating rhythm of the business. We start from the process — the queue, the approval, the handover — and place intelligence at the exact decision point where a human is currently guessing.

Every deployment ships with evaluation harnesses, human-in-the-loop review, cost telemetry and rollback. Models are swappable; the operating logic around them is ours to keep stable.

60%
manual triage removed
4x
faster case resolution
100%
decisions auditable

What we deliver

  • AI opportunity map with ROI model
  • Production retrieval pipeline
  • Agent workflows with audit trail
  • Evaluation harness and dashboards
  • Model routing and cost controls
  • Runbooks and team enablement

Working stack

PythonTypeScriptVector databasesModel gatewaysKubernetesAirflow

Opportunity mapping and AI roadmap

Process teardown across departments, scored by value, feasibility and data readiness, resolved into a sequenced build plan rather than a slide deck.

Retrieval and knowledge layers

Document ingestion, chunking, embeddings, vector search and permission-aware retrieval so answers cite your own systems of record, not the open web.

Agentic workflows

Multi-step agents with tool access, deterministic guardrails, approval gates and full audit trails for procurement, support triage and back-office operations.

Predictive and scoring models

Intent scoring, churn and propensity, demand forecasting, anomaly detection on spend — wired straight into the queues your teams already work.

Evaluation and observability

Golden datasets, regression suites, drift monitoring, latency and token-cost dashboards, plus red-team passes before anything reaches production.

Deployment and enablement

Private-cloud or VPC deployment, model routing and fallback, cost ceilings, and training so internal teams can operate the system without us.

02 — Capability

Security & technology assessments

Know exactly what you are running.

Independent audits of architecture, code, cloud and controls — delivered as a ranked remediation plan your engineers can execute, not a PDF nobody opens.

We assess systems the way an attacker and an acquirer would: what can be reached, what can be exfiltrated, what will break under load, and what will fail an enterprise procurement review.

Findings arrive severity-ranked with reproduction steps, blast radius and a fix estimate — then we stay on to help close them.

2 wks
typical assessment
100%
findings reproducible
0
surprises at audit

What we deliver

  • Severity-ranked findings register
  • Threat model and data-flow maps
  • Cloud posture report
  • Control gap and evidence matrix
  • Remediation roadmap with estimates
  • Retest and closure report

Working stack

OWASP ASVSBurp SuiteSemgrepTerraformSIEM toolingThreat modelling

Application and API security review

Authentication, authorisation, tenancy isolation, injection surfaces, secrets handling and dependency risk, tested against OWASP ASVS.

Cloud and infrastructure posture

IAM sprawl, network exposure, storage permissions, logging gaps, key management and disaster recovery readiness across AWS, Azure and GCP.

Architecture and scalability audit

Data model review, coupling and failure-domain analysis, load and cost modelling, plus a pragmatic modernisation path for legacy estates.

Code quality and technical due diligence

Maintainability, test coverage, delivery throughput and key-person risk — the diligence pack investors and acquirers ask for.

Compliance readiness

Gap analysis and control mapping for SOC 2, ISO 27001, GDPR and India DPDP, including policy scaffolding and evidence collection.

Remediation partnership

Prioritised backlog, engineering support to land the fixes, and a retest pass that closes findings on record.

03 — Capability

Enterprise resource planning systems

One ledger for the whole company.

Finance, procurement, inventory, stores and people — implemented as a single connected backbone, either on OnePay and OneHR or against your existing ERP.

We have built and run ERP for our own group, so we design around the parts most implementations get wrong: approval matrices that mirror real delegation, inventory that reconciles to the rupee, and month-end that closes without a spreadsheet rescue.

Deployments are phased by module with parallel-run periods, so operations never stop while the ledger moves.

5 days
faster month-end close
98%
three-way match rate
1
source of financial truth

What we deliver

  • Process and approval matrix design
  • Configured module rollout plan
  • Validated data migration
  • Bank and tax integrations
  • Controls and audit framework
  • Finance team training

Working stack

OnePayOneHRPostgreSQLSAP / Oracle connectorsBanking APIsPower BI

Finance and controlling

Chart of accounts design, multi-entity and multi-currency ledgers, budgeting and variance control, payment runs, reconciliation and cheque management.

Procure-to-pay

Vendor onboarding and scoring, requisitions, RFP and negotiation cycles, purchase orders, goods receipts, three-way match and dispute handling.

Inventory, stores and assets

Multi-warehouse stock, material issuance, batch and serial tracking, cycle counts, asset registers and depreciation schedules.

People operations

Org structures, joiner-mover-leaver flows, attendance, leave, expenses, payroll inputs and appraisal cycles connected to the finance ledger.

Migration and integration

Legacy data extraction, cleansing, mapping and reconciliation, plus connectors to banking, tax filing and existing SAP, Oracle or Tally estates.

Reporting and controls

Board-grade dashboards, segregation of duties, maker-checker enforcement and immutable audit logs across every transaction.

04 — Capability

Automation-backed systems development

Remove the handoffs, keep the control.

Workflow engines, integration fabric and robotic process automation that eliminate re-keying between systems while preserving approvals and audit history.

Most operational drag is not a missing feature — it is a human copying data between two systems that were never introduced. We map those seams and close them with durable, observable automation.

Every automation is idempotent, retry-safe and monitored, with a clear escalation path when a step needs a human.

70%
manual steps removed
3x
throughput per operator
<1%
exception rate

What we deliver

  • Process map with cycle-time baseline
  • Deployed workflow engine
  • Integration and connector suite
  • Document extraction pipeline
  • Exception handling playbooks
  • Automation health dashboard

Working stack

Node.jsTemporalKafka / queuesOCR pipelinesREST & GraphQLGrafana

Process discovery and mapping

Time-and-motion review of the real workflow, cycle-time baselining, and an automation candidate list ranked by hours recovered.

Workflow and rules engines

Configurable state machines, SLA timers, escalation ladders and delegation rules that business owners can adjust without a release.

Integration fabric

Event-driven pipelines, webhooks, queues, idempotent connectors and reconciliation jobs between ERP, CRM, HRMS and third-party services.

Document and data automation

OCR and intelligent extraction for invoices, POs and contracts, validation rules, and straight-through posting with exception queues.

Scheduling and orchestration

Cron and event orchestration, dependency graphs, backfills and replay, with alerting on drift, failure and latency.

Monitoring and continuous tuning

Automation health dashboards, exception rate tracking and quarterly tuning cycles so coverage grows rather than decays.

05 — Capability

Custom programming

Software built exactly to the brief.

Full-stack product engineering teams that take a problem from discovery through architecture, build, launch and long-run support.

When nothing off the shelf fits, we build. Product discovery, design systems, backend architecture, mobile and web clients, data platforms and the delivery discipline to keep shipping after launch.

Teams are senior, embedded and accountable to outcomes — with code, infrastructure and documentation owned by you from day one.

6 wks
to first production release
90%+
critical path test coverage
24/7
support coverage

What we deliver

  • Discovery and architecture pack
  • Design system and UI kit
  • Production application
  • CI/CD and infrastructure as code
  • Test suites and documentation
  • Support and SLA agreement

Working stack

TypeScriptReactNode.jsPythonPostgreSQLAWS / GCP / Azure

Product discovery and design

User research, service blueprints, interaction design and a component-driven design system that survives beyond the first release.

Web and mobile engineering

React and TypeScript front ends, React Native and native mobile, accessibility, performance budgets and offline-tolerant experiences.

Backend and platform

API design, event-driven services, multi-tenant architecture, background processing and infrastructure as code across major clouds.

Data engineering and analytics

Ingestion pipelines, warehouses, semantic layers and self-serve dashboards, with lineage and quality checks built in.

Legacy modernisation

Strangler-fig migrations off ageing stacks, incremental re-platforming and data reconciliation with zero-downtime cutovers.

Managed engineering and support

SLA-backed maintenance, on-call rotations, release management and a roadmap partner rather than a ticket queue.

06 — Capability

Media & publishing technology ecosystems

The stack behind modern publishing.

Editorial platforms, ad serving, audience data and monetisation infrastructure — built by a team that runs media businesses of its own.

OneLab grew inside a media group, so publishing technology is native ground: content pipelines, ad delivery, consent and identity, subscriptions and the revenue reporting that ties them together.

We build ecosystems, not point tools — editorial, audience, advertising and finance sharing one identity layer and one set of numbers.

22%
average yield uplift
9.4B
impressions served
<1s
median page load

What we deliver

  • Content architecture and CMS build
  • Ad server configuration and integration
  • Publisher operations tooling
  • Consent and identity layer
  • Subscription and paywall stack
  • Unified revenue reporting

Working stack

OneNetworkHeadless CMSPrebidCDP toolingCDN & edgeBigQuery

Editorial and content platforms

Headless CMS architecture, structured content models, multi-brand and multi-language publishing, workflow approvals and scheduled releases.

Ad serving and yield

Direct and programmatic delivery, header bidding, pacing and forecasting, brand safety controls and floor-price optimisation on OneNetwork.

Publisher and inventory operations

Publisher onboarding, inventory taxonomy, placement management, discrepancy reconciliation and revenue-share payouts.

Audience data and identity

First-party data collection, consent management, segmentation, cohort modelling and cookieless identity strategy.

Subscriptions and commerce

Paywall logic, entitlement services, metering, bundling, churn analytics and billing integrations.

Distribution and performance

Edge delivery, Core Web Vitals engineering, structured data and SEO architecture, syndication feeds and newsletter infrastructure.

How we engage

Diagnosis first. Then a system that stays fixed.

01

Discover

Two to three weeks inside your operation: interviews, system access, data review and a written diagnosis of what is actually costing you.

02

Architect

Target architecture, delivery sequence, integration map, commercial model and the success metrics we will be judged on.

03

Build

Fortnightly increments in your environment, demoed live, with security and performance testing built into the pipeline rather than bolted on at the end.

04

Deploy

Phased cutover with parallel run, data reconciliation, role-based training and documentation your team can actually operate from.

05

Operate

SLA-backed support, monitoring, quarterly tuning and a roadmap that keeps the system compounding in value.

Tell us what is slowing the business down.

Bring us a process, a legacy system or a blank sheet. We will come back with a diagnosis, an architecture and a delivery plan you can hold us to.