Same data. Five different jobs to do with it.
A provost defending a budget, an advisor with 400 students, an IR office facing an accreditation visit and an IT team that has been burned by an integration promise all need the same warehouse and completely different things on top of it.
Know which initiatives are actually working.
You are asked to defend a budget for programmes whose effect nobody has measured. We run matched-comparison efficacy studies each term so you can name the ones that work and stop funding the ones that do not.
- Efficacy assessment for every programme and policy, each term
- Effect per rupee, so reallocation arguments have numbers behind them
- Institution-wide KPIs against one agreed set of definitions
- Board-ready reporting that does not need re-checking
- Only 47% of the 220 initiatives in our 2026 study had a measurable effect
- Two of our own partner deployments showed no effect in year one
Prioritise outreach without guesswork.
An advisor with 400 assigned students cannot triage by intuition, however good it is. The queue tells them who to contact today, why, and what has already been tried by someone else.
- Outreach queues ranked by predicted impact, not raw risk
- Contributing factors on every student so the first call is informed
- Shared case notes across advising, faculty and student services
- Closed-loop outcome recording that feeds the next study
- −31% advisor time spent working out who to contact, measured by time-and-motion
- Duplicate outreach to the same student fell from 14% to under 3%
Find the barriers, then remove them.
Students rarely leave for one reason. They leave because a prerequisite was full, a fee hold blocked registration, and nobody noticed the assessment they missed in week three.
- Course demand forecasting by programme and cohort
- Prerequisite and sequencing bottleneck detection
- Fee-hold and registration friction analysis
- Term-to-term persistence modelling with named factors
- +7.8pt on-time completion in gateway courses at a state-affiliated cluster
- +4.1pt term-to-term persistence at a 28,000-student university
Efficacy studies that survive scrutiny.
Before-and-after comparison is not evidence, and your accreditor increasingly knows that. We use matched comparison with reported effect sizes and confidence intervals, and we publish our method so you can defend it.
- Matched-comparison design rather than pre/post
- Effect sizes with confidence intervals, not percentage-point anecdotes
- Full method documentation for accreditation and audit
- Direct warehouse access for your own analysis in SQL or R
- Every method we use is published in the research section
- Harnesses and simulated corpora released so results can be reproduced
Integration that does not become your job.
You have heard “we just need a data feed” before, and then spent a year building it. Integration is our deliverable, with named owners and a validation step where you confirm our numbers match yours.
- Connectors we build and maintain, not a spec we hand you
- Reconciliation phase where your reporting is the source of truth
- India data residency by default; regional options on request
- SSO via your identity provider; SCIM provisioning on request
- We act as processor under your instruction — see Security & data
- No cross-institution data pooling and no model training on your data for others
Most institutions are three of these at once.
Tell us the problem in your own words and we will say which part of the platform addresses it, and which part does not.