// from messy files to useful answers

Your data is in there somewhere. We’ll get it out.

We help teams make sense of complex data. Our dashboards and AI agents make everyday work easier.

selected work

Built from real data.

Client data stays private, so these are live demos on illustrative data. Try them.

Ranked by total spend

₹ PER PERSON · ILLUSTRATIVE DATA

DASHBOARD · PUBLIC FINANCE

Who funds justice?

A comparative dashboard of police, prison, and judiciary expenditure across Indian states.

0 days
MATCH r = 0.00
FIRES · PUNJAB, HARYANA AQI · DELHI

30 DAYS · ILLUSTRATIVE DATA

ANALYSIS · AIR QUALITY

How many days does smoke take?

Measuring the time lag between crop-residue fires in Punjab and Haryana and air quality in Delhi. Drag the slider until the two curves line up.

BEFORE — RAW PDF
AFTER — STRUCTURED TABLE
HeadBudget Est.Revised Est.
Police — salaries₹ 4,218 Cr₹ 4,102 Cr
Police — equipment₹ 312 Cr₹ 287 Cr
Prisons — maint.₹ 189 Cr₹ 201 Cr
Judiciary — infra₹ 540 Cr₹ 498 Cr
Judiciary — salaries₹ 1,024 Cr₹ 1,056 Cr
Legal aid₹ 78 Cr₹ 64 Cr
← DRAG →

DATA · AUTOMATION

From budget documents to structured data

Converting state budget publications into analysis-ready datasets.

Pick a month
Taxable value—
Input tax credit—
Net GST payable—

₹ LAKH · ILLUSTRATIVE DATA

REPORTING · EXCEL

Automated GST MIS reporting

A monthly reporting process reduced from hours of manual work to a single click.

services

From the file you dread opening to the dashboard you check every morning.

We work across the whole journey of your data: getting it out of PDFs and spreadsheets, shaping it into something you can see, and automating the work that keeps it current. Use one service, or let us run the full pipeline.

01

Data & insights

Scattered sources, one dataset.

We turn PDFs, websites, spreadsheets and legacy reports into one clean dataset. Every important row is checked against its source, so the numbers stand up to scrutiny.

You get

  • A clean, analysis-ready dataset in Excel or CSV
  • Documented schema and source notes
  • A short brief on the key findings
Dist. Police ...... 2055 00 109  4,218
Civil & Sess. Courts 2014-00-105 *540
Child Welfare (n.4) 2235/02/102   312
↓ extract / clean / validate
head codedescriptionstatus
2055-00-109District Police✓ matched
2014-00-105Civil & Sessions Courts✓ matched
2235-02-102Child Welfare✓ matched

A good fit if your data is spread across files that do not agree, and you need one version you can trust.

Talk about your data
02

Dashboards

The view your team checks every morning.

We design dashboards around the questions your team actually asks, with the filters and views needed for day-to-day decisions.

You get

  • An interactive dashboard with the filters you need
  • Views built around your team’s real decisions
  • A handover and a guide to keeping it updated

POLICE SPEND · ₹ PER PERSON · ILLUSTRATIVE DATA

A good fit if your team rebuilds the same report every month or waits for someone to pull numbers by hand.

Plan a dashboard
03

AI agents

The routine work, done while you sleep.

Custom AI agents extract data, generate reports and deliver scheduled summaries—with human review built in before anything goes out.

You get

  • An agent built around your workflow
  • Scheduled reports and summaries, on autopilot
  • A human review step before anything goes out
agent / weekly_summaryneeds review
  1. ✓read new documents3 files
  2. ✓extracted and validated rows142 rows
  3. ✓drafted summary report1 page
  4. ◦waiting for your review

A good fit if your team spends hours copying numbers from documents or writing the same update each week.

Automate a task

How a project runs.

Every engagement follows the same four steps, whichever service you start with.

  1. step 1

    Scope call

    We look at your files and agree on what “done” means.

  2. step 2

    Sample

    A small slice of the output first, so you can check the format.

  3. step 3

    Build

    The full dataset, dashboard or agent, with regular check-ins.

  4. step 4

    Handover

    Files, documentation and a walkthrough for your team.

about

Who we are.

Most organisations already have the data they need. It’s just trapped in PDFs, scattered across spreadsheets, and buried in reports no one has time to read. Tell us what you want. We’ll go and get it.

yjt labs is led by Yaj. The firm exists to make inaccessible data usable: structured, analysed, and put to work.

Founder
Yaj
Focus
Data, dashboards & AI
Approach
Source-first

Questions, answered.

Something else on your mind? Ask us directly.

What kinds of files can you work with?

Text PDFs, scanned PDFs, Excel and CSV files, web pages and older report formats. If it holds numbers, we can usually get them out.

How do you make sure the numbers are right?

We check extracted figures and totals back against the source documents and keep descriptions matching the original text, so anyone can trace a number back to where it came from.

Do you work with government and public data?

Yes. Much of our work so far has been structuring Indian state budget documents across multiple years into a single, comparable format.

Who owns the data and the work?

You do. Everything we produce for your project is handed over to you at the end.

How long does it take, and what does it cost?

Both depend on the project, so we don’t quote blind. Send us a note about your data, and after a short scope call we’ll come back with a timeline and a price.

How do we get started?

Send us a short note about your data and what you want to do with it. We’ll reply with questions and set up a scope call.

contact

Tell us about your data.

A few lines is enough. What do you have, and what do you need it to do?

We use your details only to reply to you, and we don’t share them. Privacy