Dilyas Studio / Independent AI & workflow consulting

Less admin between you and the work that matters.

I help small service businesses simplify the work between inboxes, forms, and spreadsheets. Together, we pick one recurring bottleneck and build a practical improvement around how your business runs.

Work directly with Biswajit Sandilya, founder and independent consultant.

One workflow to startScope agreed before the buildHuman review where it matters

Where does the work keep getting stuck?

The best starting point is a task your team repeats often enough to recognize the cost.

Chasing details before work can start

Gather missing information, flag incomplete requests, and prepare a useful summary before someone schedules or quotes.

Copying the same information twice

Explore ways to move approved details between your existing tools, after checking what those tools actually support.

Keeping track of what happens next

Make outstanding estimates, internal tasks, and customer follow-ups easier to see, assign, and act on.

Illustrative example / fictional request, not a client result

Turn a scattered request into a clear next action.

A small service firm receives a request that is missing key details. Here is the kind of workflow we could design together.

What arrives in the inbox

Subject: Estimate for the office

"Hi, we need three rooms painted before our team moves in next month. Could someone send a price? I can send photos later."

Someone still needs to clarify the address, room sizes, condition, and timing.

What the workflow prepares

Request
Office painting, three rooms
Timing
Next month; exact date needed
Missing
Location, dimensions, photos
Next action
Draft a request for the missing details
Approval
Owner checks the draft before sending

What we would test: accuracy of the summary, time spent preparing the reply, and how often a person needs to correct it. Pricing and scheduling stay with your team.

Proof of work / fictional customer-signal pilot

From support noise to executive-ready product signal.

This internal-tool demo shows how Dilyas Studio works through an ambiguous operating problem: define the pain, analyze the data, separate symptoms from root cause, and package the evidence so a leadership team can decide what to do next.

What the pilot does

The dashboard reads support tickets, account data, and Jira timing to surface recurring customer patterns, quantify business impact, and route validated signals through a human review workflow.

369matching tickets analyzed
$15.3Mgross ARR exposure surfaced
3executive-ready priority signals

Why it matters

The work demonstrates an evidence-first consulting approach: customer pain is visible before the solution is obvious, and the right next step is often better workflow, ownership, and decision support.

Start with one problem worth solving.

You bring a recurring task and the tools you use today. I help work out whether a small change, a straightforward automation, or AI is a sensible fit.

Scope, fees, timing, software costs, and ongoing support are agreed before implementation.

  1. A 20 or 30 minute conversation

    Talk through what happens today, where time goes, and what a better result would look like. Decide whether a closer review makes sense.

  2. A focused workflow review

    Map the steps, check tool access, identify exceptions, and propose one improvement with a clear way to assess it.

  3. A small, agreed pilot

    Test representative examples with you, review mistakes, and document how the workflow is used and paused. Decide what happens next from the results.

Biswajit Sandilya, founder of Dilyas Studio

Biswajit Sandilya

Founder, Dilyas Studio

The person you talk to is the person doing the work.

I run Dilyas Studio as an independent AI consultancy. You work directly with me to understand the problem, make implementation decisions, and review what is built.

For a first project, I propose a limited scope and an agreed test. That gives us something concrete to evaluate before considering a broader rollout.

A good fit: a small service business with a recurring administrative task, access to its current tools, and someone available to test the change with me.

Before we start.

Does every project need AI?

No. Rules, templates, or a simpler process may be enough. AI is useful where interpreting or drafting information helps, with its output checked against the task.

Will this work with our existing software?

That needs checking. Available integrations, permissions, and subscription levels affect what is possible. The review identifies these constraints before a build is proposed.

Will it send messages without our approval?

We agree which actions need review. For the example above, the reply stays a draft until a person approves it. Exceptions and a way to pause the workflow are part of the design.

What happens after the pilot?

We compare the work against the agreed criteria and decide whether to refine, continue, or stop. Ownership, documentation, and any maintenance arrangement belong in the project scope.

What's the task you keep putting off?

Tell me what repeats, which tools are involved, and where it gets stuck. That is enough to start a useful conversation.

Book a 20 or 30 minute conversation, or send the details by email.

This opens a prefilled email to Dilyas Studio so you can review it before sending.