Software for education and EdTech
An education platform has two days a year that decide its reputation: the day admissions open and the day results are published. On both, every user arrives within the same twenty minutes.
Every sector page here leads with the modelling problem that sector actually has — the exception that breaks the schema, the number two systems disagree about. If it could be rewritten for another industry by swapping six nouns, it would not be on this site.
The load is a spike, not a curve
Most systems are sized for average load with headroom. An education system is sized for a spike two orders of magnitude above its average, lasting under an hour, on a date everyone knows in advance. A platform that performs beautifully for eleven months and falls over on results day has failed, and the failure is public and memorable.
The design responses are unglamorous: the results page is a static file on a CDN rather than a query, the admissions form queues writes rather than contending on the same rows, and the whole thing is load-tested against the real peak before the date rather than discovered on it. Knowing the date in advance is an enormous advantage and it is routinely wasted.
Academic structure resists clean modelling
A student belongs to a batch, which belongs to a programme, which has a curriculum that changed in the third year for students admitted after a certain date. Some students are repeating a subject from a previous curriculum. One has transferred credit from another institution. Two are on an approved variation because of a scheduling conflict.
These are not edge cases, they are the normal state of any institution more than a few years old, and a schema that assumes a clean cohort will need manual overrides within a year. Modelling curriculum as versioned, and a student’s programme as an enrolment with its own effective dates, is the arrangement that survives.
Assessment integrity
Online assessment introduces problems that do not exist on paper: whether the person answering is the person enrolled, whether the connection dropping should void an attempt, and what happens when two hundred students all lose connectivity because a regional network had a bad afternoon.
Proctoring is a partial answer with real privacy costs that deserve a deliberate decision. Assessment design is often the better one — questions that draw from a large bank, are time-boxed sensibly, and reward understanding over recall are harder to game and less dependent on surveillance.
The six things that decide a education & edtech project
These are where the schedule goes, and where a generic implementation quietly fails.
Intake and results spikes
Sized and load-tested for the twenty minutes that matter, on a date everybody already knows.
Curriculum versioning
Programmes that changed mid-cohort, transfer credit and approved variations — modelled rather than overridden by hand.
Admissions workflow
Application, document verification, merit lists, seat allocation and fee payment, with an auditable decision trail.
Learning delivery
Content that works on a shared phone over a slow connection, because that is what a large share of students have.
Assessment integrity
Question banks, sensible time-boxing and a considered position on proctoring and its privacy cost.
Fee management
Instalments, concessions, scholarships and late fees resolved by a rule, with receipts that reconcile to the ledger.
Systems for education & edtech
Not a product list — these are the shapes of system this sector keeps needing, built to fit rather than configured to approximate.
- Admissions portals with document verification and merit-list generation
- Student information systems with versioned curriculum and enrolment history
- Learning management and content delivery built for slow connections
- Assessment platforms with question banks and integrity controls
- Fee management with instalments, concessions and ledger reconciliation
- Attendance capture — biometric, RFID or app-based
- Parent and student portals with role-appropriate visibility
Education & EdTech, answered
The questions that come up on a first call in this sector specifically.
Can the system handle admissions day traffic?
It can if it is designed for the spike rather than the average, which is entirely feasible because you know the date months ahead. In practice that means serving read-heavy pages such as results and merit lists as static files from a CDN rather than as database queries, queueing writes on the application form so thousands of submissions do not contend on the same rows, and load-testing against a realistic peak weeks before the date. The advantage of a predictable spike is routinely wasted, and it is the cheapest kind of scaling problem to solve.
Should we build or buy a learning management system?
Buy, in most cases. Moodle and the commercial platforms are mature, and rebuilding course delivery, enrolment and gradebook functionality is rarely a good use of an institution’s budget. Building makes sense where your pedagogy or your assessment model is genuinely unusual, or where the integration burden with your student information system is so heavy that the LMS is only a shell anyway. We will say which situation you are in, including when the answer is that you need integration work rather than a new platform.
How do you handle students on poor connections?
By treating it as the default rather than the exception. Content is served at multiple qualities with the low end genuinely usable, pages work without heavy JavaScript, downloads resume rather than restart, and anything a student has opened once is available offline where the platform permits it. We test on throttled connections and mid-range Android phones as part of the normal test matrix, because a platform that only performs on good hardware has excluded a large share of the students it was built for.
What about online exam integrity?
There is no complete answer, and it is worth being direct about that rather than selling proctoring as one. Camera-based proctoring catches some behaviour, carries real privacy costs, and generates false accusations that are unpleasant to adjudicate — it deserves a deliberate institutional decision rather than a default. Assessment design often does more: large randomised question banks, sensible time-boxing, and questions that reward understanding over recall are harder to game and do not require surveillance. We will build either, and we will tell you what each actually achieves.
What this sector usually needs from us
- BuildWeb Application DevelopmentDashboards, portals and internal tools that hold up under real data volumes — not a prototype that falls over at ten thousand rows.Read more
- BuildCustom Software DevelopmentCustom platforms, marketplaces and SaaS. We write the parts that are specific to your business and buy the parts that aren’t.Read more
- GrowUI/UX DesignProduct design and design systems. Usually the first thing that turns out to be wrong, and the cheapest thing to fix.Read more
- RunCloud & DevOpsDeploys that are boring on purpose. Infrastructure as code, real environments, and a bill you can read line by line.Read more
Working in education & edtech?
Tell us what your team works around every week. That is usually where the real project is, and it is a better opening than a requirements document.