How I work
How I work, and what decides the cost.
A project starts with a first call. The usual next step is a Workflow & Data Audit, which ends in a written recommendation you can stop after. Five things decide the cost: integrations, data readiness, platforms, security needs and who runs it afterwards. I lead the consulting on every project, and YourTechConsultants delivers the work.
A first call, an Audit, a pilot, a build.
I start with a first call. The usual next step is a Workflow & Data Audit. It ends with a written recommendation, and you can stop there. If the right answer is to do nothing, the recommendation says so.
First call
A conversation about the problem, the systems and the data. The first call is free.
Yes, no, or a referral.
Workflow & Data Audit
Fixed scope: baseline the process, check data readiness, set one measurable success criterion. You get a written recommendation, including “do nothing” where that is right. The Workflow & Data Audit is paid.
You can stop after the written plan.
Pilot
One success criterion measured against the Audit baseline: a thin working slice on real data and a short evaluation note.
Go or no-go.
Build
Built in milestones, each with a fixed scope.
A pilot and a build come after the Audit, and only if you choose to go on.
I design it with you. YourTechConsultants delivers it.
I lead the consulting on every project. You work with me on the problem and the design, and YourTechConsultants delivers it.
I understand the problem
You describe the workflow or decision that costs you time. I ask about the systems, the data and who does the work today.
I design the system
I work out what to build or change, which tools fit and what to leave out. The recommendation puts the design in writing.
YourTechConsultants delivers it
The pilot tests the design on real data, and the build then follows in milestones.
What stays the same, what varies, what I need.
What stays the same
- I lead the consulting, and YourTechConsultants delivers the work.
- The Audit ends in a written recommendation, and you can stop after it.
- The recommendation sets out what I would build or change, what I would leave out and the risks to manage.
- Nothing goes further until you say so.
What varies
- The cost, through the five drivers below.
- The tools. I choose the lightest one that does the job, and the recommendation says why.
- What I find when I look at your systems and data.
- The size of the pilot and the build, which follows the problem.
What I need from you
- The workflow or decision, in plain words.
- A way to read the data and see the systems involved.
- Someone who knows what the data means.
- Someone who decides, and someone who checks the output.
Five things decide the cost.
The same request can be small in one business and large in another. These five things explain the difference, and I ask about each in the first call. If you have a budget range in mind, say so. It helps me suggest a first step that fits.
- Driver 01
Integrations
A system often has to connect to others, such as WhatsApp, an ERP, a SQL database or a website. Each has its own way in and its own rules. More systems, and less cooperative ones, mean more work.
- Driver 02
Data readiness
Data that is complete, consistent and reachable is quick to use. Data that is scattered or duplicated has to be sorted out first. The Audit checks it before I recommend anything, and the recommendation starts from what your data can support.
- Driver 03
Platforms
A web app, an Android app and a Google Play listing are different amounts of work. So is each extra platform.
- Driver 04
Security needs
Who may see which data, where it may be stored and what must be recorded. The more a system has to protect, the more care and checking it needs.
- Driver 05
Who runs it afterwards
You, your team or someone else. That changes both the design and the cost, so I raise it in the recommendation.
Running costs set by the platforms
Some costs belong to the platforms your system runs on, such as message fees on WhatsApp, hosting, store accounts and usage charges from a language-model provider. The recommendation lists the ones that apply, so they are visible before you decide.
What the recommendation covers for an AI feature.
A language model writes fluent answers, and fluent is not the same as correct. So for any AI feature, the recommendation sets out how the feature will be checked before it goes live, in six parts.
What a good answer is
You and I agree what a correct answer looks like in your work, using real examples.
Real cases
Real questions or documents from your workflow, shared with your permission, with the right answers marked by you.
A test run
The feature runs on those cases, and each answer is read for errors, vague wording and too much confidence.
The awkward cases
Vague questions, missing information, and requests the feature should decline.
A person where a wrong answer costs something
A person checks the output before anything acts on it. Totals, tax and stock counts stay in ordinary code.
How success is judged
What is checked, who judges the result and what happens when the feature gets an answer wrong.
Data and models: what the recommendation states.
For any system that uses a language model or handles your data, the recommendation states three things before you commit.
- Which model provider and data flow the system uses.
- How long data is kept.
- Whether it is used to train a model.
Each is checked against the provider’s current terms before it is written down.
Please keep client data out of your first message. Read how I handle what you send.
Send the twelve questions first.
The scope brief is a short sheet of twelve questions about your workflow, systems, data, security, timing and budget. Fill it in yourself, forward it to a colleague who knows the answers, or send it to me.
Questions before you write.
Can I see the work before I contact you?
Yes. The Work page has the case studies, and ArPar Private Browser is live on Google Play. Much of YTC’s work is for private clients, so it is shown here with client details removed.
Do I need to know what I want before I write?
No. A description of the problem is enough, for example “orders arrive on WhatsApp and I retype them”. I will say what I think the first step could be.
Who will I work with?
You work with me, Raghav Palriwala, the consulting lead. I work with you to understand the problem and design the system, and YourTechConsultants delivers it.
Where are you based?
YourTechConsultants is based in Indore, India (IST) and works remotely with clients worldwide.
Which tools do you use?
That depends on the work. The services overview links to the four service pages, which list the tools for each kind of work. The recommendation says which fits yours and why.
How do I send you the details?
Use the scope brief, the contact form, or email yourtechconsultants24x7@gmail.com. I read every message and reply personally.
Printed from yourtechconsultants.com/how-i-work/. Addresses in brackets are pages on that site. To reach me, email yourtechconsultants24x7@gmail.com.
Tell me the workflow you want gone.
I read every message and reply personally.
Write to me
Two or three sentences is plenty: the workflow or decision that costs you time, and the tools you already use. No client data, please.