Skip to main content
Chadura
Consultation

Start small, with the price on the page

Each of these is a short, fixed-fee engagement with the scope written down and the deliverables handed over as something you own. One budget holder can approve them without a procurement cycle, and the fee is credited in full if you decide to carry on with the larger work behind it.

SGD 9,800 2 weeks

Cloud Cost & Architecture Review

A two-week, fixed-fee review of your AWS or Google Cloud spend and architecture. You get written findings, a savings model costed in SGD, and a remediation plan ordered by payback. No tooling to buy, no retainer, no implementation commitment.

Scope and deliverables
SGD 12,500 3 weeks

Agentic Automation Pilot

Three weeks to put a single real workflow into production on self-hosted n8n, with language models doing the judgement work and a human approving anything consequential. Running on your infrastructure, so your data never leaves your tenancy.

Scope and deliverables
SGD 8,500 2 weeks

Lead Engine & CRM Setup

Two weeks to put a working lead pipeline behind your website: spam filtered out, every enquiry scored and routed, a self-hosted CRM that records which channel produced it, and a booking link so buyers can skip the wait. We built this for ourselves first.

Scope and deliverables

The wider engineering work

Not everything fits a fixed fee. These are the engagements we are usually brought in for, with enough detail to tell whether yours is one of them. Each says how the work normally starts, because that is the part most service pages leave out.

Core systems engineering

High-performance computing

Python, Rust and Julia development

We write the parts of a system where the general-purpose stack has run out of room: a hot path that has to hold a latency budget, a service that cannot afford a garbage-collection pause, a simulation that has to finish overnight rather than over a weekend. Usually this means Rust at the boundary, Python where the team already lives, and Julia where the maths is the product.

How it usually starts

Most of this work starts as a rewrite of one component rather than a platform project. We profile what you have, agree the budget the component has to meet, and replace it behind the existing interface so the rest of the system does not move.

  • Rust
  • Tokio
  • Python 3.13
  • Julia
  • Zero-copy IPC
  • PyO3
Applied intelligence

AI and ML transformation

The useful version of this is rarely a model. It is the pipeline around one: how training data is collected and labelled, how a model gets evaluated against the decision it is actually replacing, and what happens when it is wrong. We build that pipeline, and we are equally willing to tell you the problem does not need a model.

How it usually starts

Starts with a single decision your team currently makes by hand and a measurement of how well it is made today. Without that baseline there is no way to say whether the model helped.

There is a fixed-fee way to start this
  • PyTorch
  • Private LLM fine-tuning
  • Vector embeddings
  • Evaluation harnesses
  • CUDA
Distributed streaming

Big data and streaming analytics

Most data problems presented to us are not volume problems; they are trust problems. The numbers in two dashboards disagree and nobody can say which is right. We build pipelines where lineage is traceable end to end, late and duplicate events are handled explicitly, and the definition of each metric lives in one place.

How it usually starts

We usually begin by reconciling two figures that should match and do not. That exercise surfaces the real state of the pipeline faster than any audit document.

  • Apache Kafka
  • Apache Spark
  • Lakehouse ETL
  • dbt
  • Event telemetry
Embedded and edge

Systems and IoT engineering

Hardware that has to run unattended for years, on a power budget, over a link that drops. The engineering is mostly in the failure cases: what the device does when the network is gone, how it is updated in the field without bricking a fleet, and how you find out something is wrong before a site visit tells you.

How it usually starts

Typically a firmware and protocol review of an existing fleet, or a bring-up for a new board where the telemetry design is settled before the first unit ships.

  • Embedded firmware
  • MQTT
  • Industrial sensor mesh
  • OTA update paths
  • Edge compute

Cloud and operational scale

Cloud modernization

Cloud migration and modernization

Lifting a workload into a cloud account is the easy half, and the half that makes the bill worse. The work that matters is deciding what to move, what to rewrite, and what to switch off, then sequencing it so the business keeps running throughout. We do the sequencing in writing before anything moves.

How it usually starts

Almost always starts with the cost and architecture review below, because the migration plan depends on numbers most teams do not have yet.

There is a fixed-fee way to start this
  • AWS
  • Google Cloud
  • Azure
  • Kubernetes
  • Terraform
  • Docker
Platform engineering

IT infrastructure and platform engineering

The platform layer your product team should not have to think about: networks, identity, secrets, backups that have been restored at least once, and a path to production that does not depend on one person's laptop. Defined as code, so the state of your estate is something you can read rather than something you have to ask about.

How it usually starts

Often a remediation engagement after an audit or an incident, scoped against the specific findings rather than sold as a platform rebuild.

  • Terraform
  • Enterprise Linux
  • Zero-trust networking
  • Secrets management
  • Backup and restore drills
Engineering delivery

Offshore engineering teams

A squad that reports into your planning process rather than a separate account structure, with a technical lead who is accountable for what ships. Our working day overlaps the Singapore and wider APAC afternoon, so review and handover happen inside one working day rather than across two.

How it usually starts

Starts at two to three engineers on a defined piece of scope, with a review at the end of the first month where either side can stop. Scaling up happens after that review, not before it.

  • Senior engineers
  • CI/CD quality gates
  • Overlapping hours
  • Defined SLAs
Institutional governance

E-filing and statutory compliance

Indian statutory filing, automated where the rules are stable and checked by a person where they are not. This sits beside our own eFiling360 product: the product covers the standard path, and this service covers the organizations whose structure does not fit it.

How it usually starts

Scoped against your filing calendar and entity structure. We will say up front which parts can be automated safely and which should stay manual.

  • GST automation
  • PF and ESI computation
  • MCA filings
  • Audit trails

Not sure which one

Book a call and describe the problem. If none of these fit, we will say so. If the right answer is a larger engagement, we will tell you that too rather than sell you a pilot that does not address it.

Book a 30-minute call WhatsApp

An engineer replies within one business day.

M

Maya Solutions Architect

Ask about products & scoping →

WhatsApp Direct Help