×
1,000 rewards. 20 cities.
No black box.
Senoa is an events and campaign platform for the moment demand outstrips supply — and someone has to explain why the answer was fair.
When an event stops being a room and becomes a queue
Mokshacreations is a product studio that builds software from first sketch to shipped release. Senoa began as event management — publishing an event, taking registrations, talking to guests, checking them in on the day.
The hard part showed up later. Organizers kept running campaigns where far more people applied than there were spots or rewards: 1,842 applications for 1,000 coupons spread across 20 cities, with per-city quotas and eligibility rules layered on top. That work was being done in spreadsheets, by hand, under deadline.
A spreadsheet can produce a list of winners. What it cannot do is survive the question that always follows — why not me? Without a record of how each decision was reached, an organizer is left saying “trust us,” which is exactly what a sponsor putting up a rewards budget will not accept.
Fair by design, and provable after the fact
We built Senoa around a single principle: every choice leaves a trail. Events and campaigns feed one selection engine, and that engine writes down everything it did.
One system for every event format
Public, unlisted, or invite-only. In-person, virtual, or hybrid. Organizers publish the event, then run registration, guest communication, and day-of check-in from the same place — instead of stitching a form builder to a mailer to a spreadsheet at the door.
Campaigns: applications with rules
A campaign is what an event becomes when demand outstrips supply. Applicants apply, organizer-defined eligibility rules decide who qualifies, and rewards are distributed against whatever criteria the campaign sets — so the shape of the campaign lives in configuration, not in a developer ticket.
A deterministic selection engine
The heart of the build. Quota-based, weighted, random, and scored selection run through one engine on a recorded seed and a frozen applicant snapshot. Re-run a draw six months later with the same inputs and you get the same 1,000 winners — the property that makes a result defensible rather than merely plausible.
An audit trail anyone can read
Every rule evaluation, quota decrement, and tie-break is written to an immutable decision record. Claude reads that record and turns it into a plain-language explanation of why a specific applicant was accepted or rejected — the answer an organizer needs when someone asks “why not me?”
The model at the edges, never in the draw
A selection platform earns trust by being predictable. That shaped where the model was allowed to sit: on either side of the decision, and never inside it.
It explains; it does not decide
Selection stays in deterministic code, where it can be replayed and verified. Claude is scoped to reading the decision record and writing it back in plain language. A model that quietly changed an outcome would destroy the guarantee the product is built on, so it was never given the chance to.
Plain-language rules, structured safely
Organizers describe eligibility the way they think about it — “students in tier-2 cities who applied before the 14th.” Claude drafts that into typed, reviewable rule objects that the organizer confirms before a campaign opens. Nothing runs until a human has seen the structured version.
Calibrated when the rule is ambiguous
On genuinely underspecified criteria, Claude asks rather than guessing at intent. For a system whose entire value is defensibility, a model that surfaces ambiguity before the draw is worth more than one that resolves it confidently and wrongly.
“The hard part was never picking a thousand names. It was being able to sit across from someone six months later and show them exactly how their application was handled. Senoa does that, and it does it without asking anyone to take our word for it.”
What it added up to
Organizer impact
Apply, qualify, select. Each stage is countable, explainable, and reproducible from the campaign record.
Per-city quotas enforced inside a single selection pass, instead of twenty spreadsheets reconciled by hand.
The selection logic left the analyst laptop it used to live on and became a versioned, tested part of the product.
Business impact
New campaign shapes are set up by organizers rather than shipped by engineers, so Mokshacreations can take on campaigns that would previously have needed custom work.
Senoa can answer a challenge to any individual decision — the difference between a platform a sponsor tolerates and one they will hand a rewards budget to.
Building something that has to be defensible?
We help teams ship AI systems where the reasoning has to hold up to scrutiny — from deterministic decision engines to audit trails a person can actually read. Let's talk.
Contact Us