Tattva in education

Universities run on processes designed for smaller intakes. We help institutions redesign those processes and build the software that runs them, then look after it through the year.

Applications read, scored and sorted in one pass, with the shortlist in gold. An illustration with sample data.

Case study

Admissions at a leading private university in India

Client
A leading private university in India
Worked with
University leadership
Scope
Admissions, from application to shortlist
Before: manual reviewAbout 2 months
After: AI-assisted processA few hours
Time to process an admissions cycle. Bars drawn to scale.

About 2 months to a few hoursTurnaround for an admissions cycle

Thousands of applicationsHandled each cycle by the same team

No additional hiresGrowth in applications without growth in headcount

The problem

Every admissions cycle brought thousands of applications, and every one of them was reviewed by hand.

Each application had to be screened against the university's criteria, each essay read and scored, and each set of documents checked. A full cycle took around two months, and the only way to take on more applicants was to hire more people.

The university's leadership wanted to handle more applications without growing the admissions team, and to score every applicant objectively.

What we did

We worked with the university's leadership to redesign the admissions process from first application to shortlist, then built an AI-assisted system to run it.

A simplified view of applications moving through the system. Gold dots have been scored, circled dots are applicants who may need support, and faded dots leave the process. The numbered stages are described below.
Shortlisting
Screens every application against the university's admission criteria and produces a shortlist.
Essay evaluation
Reads each applicant's essay and evaluates it against defined criteria.
Objective scoring
Scores every applicant on the same criteria, in the same way.
Support needs
Identifies where applicants may need support, based on their background, so the university can plan for it early.
Data verification
Checks the information applicants submit for gaps and inconsistencies before decisions are made.

The engagement followed the same process we use for every client: discovery with the people doing the work, a signed requirements document, AI-assisted development, manual and automated testing, and sign-off before launch. See how an engagement runs.

What changed

A process that took about two months now takes a few hours.

The university now processes thousands of applications in each cycle without adding a single person to the admissions team. Every applicant is scored on the same criteria, and the support an applicant may need is identified during review.

What the university said

In progress

Student performance management, with another leading private university in India.

We are in discussions to build an AI-enabled system that helps the university track how students are doing and step in earlier when someone needs support.

Where else this applies

Most institutions have processes with the same shape: high volume, careful judgement, and a team that can only read so much.

  • AdmissionsScreening, essay and interview evaluation, document checks
  • Student performanceTracking progress across courses and flagging students who need help
  • Academic advisingPreparing advisers with a clear picture of each student before they meet
  • Accreditation reportingGathering evidence and data from across departments into one place
  • PlacementsMatching students to opportunities and tracking outcomes

Explore Tattva

Everything we build is delivered through the same process and looked after under the same annual package.

  • KayalWorkflow automation that works like a spreadsheet
  • LakshyaGoal management your organisation defines
  • HandloomA fitted edition of the enterprise software you have outgrown
  • How we workSeven stages, from discovery to a year of care

Running a process at your institution that has outgrown its team?

Tell us how it works today, and we will tell you whether software would help.

Book a conversation