Hoshang Sheth logoHoshang Sheth
Pricing

What each service actually costs.

Published ranges, not "contact us for a quote". Where a project lands inside its range depends on scope, and you get a single fixed number after the discovery call - before any work starts.

All Five Services

Side by side.

Custom Software & Apps
₹30,000 - ₹60,0008-15 weeks

Your workflow does not fit any off-the-shelf tool

Drives cost: User roles, integrations, payment handling

AI Automation Systems
₹35,000 - ₹70,0005-8 weeks

A repeatable manual process is eating your team's hours

Drives cost: Number of workflows, systems connected, edge-case handling

Gen AI Applications
₹45,000 - ₹80,0006-10 weeks

Your team drafts the same content by hand, or you need code, images, or video produced at scale

Drives cost: Content volume, brand-voice complexity, review tooling

RAG & AI Agents
₹55,000 - ₹90,0008-12 weeks

Institutional knowledge is scattered and nobody can find answers

Drives cost: Knowledge sources, access control, agent actions

Machine Learning Systems
₹60,000 - ₹95,000Custom estimate

You need to see a problem coming before it costs you

Drives cost: Data quality and volume, model complexity, monitoring

Every project is quoted as a single fixed price once scoped. The ranges above reflect how much project size varies between clients - not variable billing within a project.

How It's Calculated

Why one project costs more than another.

Price tracks what it takes to ship a correct version of the system, not how many hours it takes to write. These are the factors that move a project up or down inside its range.

Scope depth

How many distinct workflows or modules the system has to cover. One well-defined process costs less than five interconnected ones.

Integrations

Every external system - your CRM, ERP, WhatsApp, payment gateway - adds authentication, error handling, and a failure mode that has to be tested.

User roles and access control

A single-user tool is straightforward. Multiple roles with different permissions multiply the testing surface, and in RAG systems they change what the AI is allowed to answer.

AI complexity

Retrieving an answer is simpler than reasoning over it, which is simpler than an agent taking actions on your behalf. Each step up adds build and verification work.

Production hardening

Monitoring, audit trails, exception queues, and retraining pipelines are what separate a system that runs unattended from a demo. They are priced in, not added later.

Data readiness

Mostly relevant to ML. Clean, well-labelled historical data shortens the work considerably. Messy or thin data means preparation before any modelling starts.

Included Every Time

What the price always covers.

  • A fixed price agreed before any work begins
  • Full source code ownership on handover
  • Deployment documentation and a recorded walkthrough
  • 30 days of post-launch support, included
  • A live staging environment throughout the build
  • No monthly fees and no vendor lock-in
Payment Terms
50%
on approval of the proposal, to begin work
20%
at working prototype
30%
before final handover

No hourly billing. No monthly retainer. Anything outside the agreed scope is quoted separately and starts only once you approve it.

Not Sure Which Fits?

Bring the problem rather than the solution. If a simpler tool or a better process would solve it faster than anything I build, I will tell you that on the call.

Book a Discovery Call
Next Step

Know roughly what you need? Let's scope it.

Book a free 30-60 min discovery call. You'll get a fixed-price proposal within 5 business days.