SJIT CHENNAI · SEPTEMBER 30th – OCTOBER 1st · 2 Days to build 1 Day to polish, record and submit
� SEPTEMBER 30th - OCTOBER 2nd · ALL ENTRANCE LEVEL 1 & 2 + ADVANCED
Build the AI
Four levels. Eight tracks one platform. Students build inside BASIX.MARKET as a live �MVP Environment — on SingularityNET MeTTa and the Hyperon AGI stack. +OMEGA AI AGENT� What they ship has to be sellable, not submittable.
4 8
TRACKS PROBLEMS
11 5
WINNING TEAMS TEAMS ADVANCE
$1,700+
IN CASH PRIZES
OMNIVERSITY
Student Registration
BASIX.MARKET · HACKATHON TRACK
02
TRACK POSITIONING
This is not a
classroom exercise.
Students work inside BASIX.Market as a live MVP environment, and what they build has to be sellable. The test is whether a university, a business or a sponsor would pay for it — or sign up because of it.�
The platform is the constraint
Not the tooling. Teams may use Omega, MeTTa, both or neither.
Real schools, real students
Activated campuses, onboarded cohorts, tracked submissions.
A buyer on the other end
Every brief ends in someone who would open it, pay for it or renew.
Student Registration
https://basix.market/lms/hackathons
BASIX.MARKET · HACKATHON TRACK
03
THE EVENT AT A GLANCE
Two days. Three tracks.
Eleven winning teams.
2
DAYS
3
TRACKS
11
WINNING TEAMS
$1,500+
IN CASH PRIZES
11
PARTNERS
Hosted at SJIT Chennai with developers from iCog Labs and the BASIX.Market Omniversity team, alongside Kenyatta University Kenya — filmed on site by BeyondtheCode.ai. Combined Virtual Sessions With Virtual Judges
BASIX.MARKET · HACKATHON TRACK
04
WHO IS IN THE ROOM
Staffed like a delivery team.
Five organisations, each with a job on the weekend — not a logo wall.
SJIT Chennai
Host campus. Students, space and the cohort we activate.
iCog Labs
Developers on the floor, mentoring teams on the AGI stack.
BASIX.Market
Omniversity
Delivery team running assessment, onboarding and review.
Kenyatta University
Kenya
Partner university joining the cohort from East Africa.
BeyondtheCode.ai
Filming the event — the footage is the proof asset.
BASIX.MARKET · HACKATHON TRACK
05
PARTNERS & ECOSYSTEM
The names behind the weekend.
Every logo here has a role on the ground or a stake in the stack. Each one links to its site.
DELIVERED BY BASIX.MARKET OMNIVERSITY · DEVELOPERS ON THE GROUND FROM ICOG LABS
SJIT · HOST CAMPUS (CHENNAI)
ECOSYSTEM & DELIVERY
BGI LABS
AFRICAN BLOCKCHAIN FESTIVAL
IGOC LABS
CODE SLAYER
HARLEM CLX
BASIX.MARKET · HACKATHON TRACK
06
ENTRANCE LEVEL 1 & 2
Build the AI Infrastructure Layer.
Three tracks. A team picks one and ships against a real brief.
TRACK 01
BASIX Marketplace
Build something sellable inside the live MVP.
Real platform data
A real buyer at the end
Any tooling, or none
TRACK 02
SingularityNET MeTTa
MeTTa as a graph relationship database.
Entities & relationships
Multi-hop or rule-based
Show the path
TRACK 03
Omega AI Agents
Stateful agents with auditable reasoning.
State across sessions
Auditable decisions
Memory that persists
BASIX.MARKET · HACKATHON TRACK
07
TRACK 01 · BASIX MARKETPLACE
What you build
should be sellable.
You are working inside BASIX.Market as a live MVP environment. The test is whether a university, a business or a sponsor would pay for it, or sign up because of it. You may use Omega, MeTTa, both, or neither — the platform is the constraint, not the tooling.
HARD REQUIREMENT
Runs against BASIX.Market data — students, cohorts, badges, quizzes and XP, IP projects, universities, marketplace listings — and produces something with a real buyer or user on the other end.
BUILD YOUR OWN IF IT
It touches real platform entities.
It produces something a student, mentor, university, business or sponsor would open.
It moves someone closer to signing up, paying or renewing.
BASIX.MARKET · HACKATHON TRACK
08
01
TRACK 01 · EXAMPLE IDEAS
Five things worth building.
Skill-to-Brief Marketplace Router
Businesses post short briefs. The system ranks students and teams against them using verified credentials and completed IP projects, and explains the near-misses as well as the matches.
University Activation Pack
Qualify an inbound school, generate its co-branded onboarding material, and show where each institution sits in the activation funnel. Sales tooling for the people selling the platform.
Student Portfolio Storefront
Turn a student's badges, quiz history and IP project shares into a public, verifiable page an employer or client would actually read.
Cohort Health Agent
Track quiz progress, XP and challenge completion, flag students drifting toward drop-off, and draft the mentor's outreach message.
Fractional CTO Workflow
Intake interview, then requirements, then build-versus-buy vendor evaluation, then an MVP roadmap and a development brief a contractor could quote against.
BASIX.MARKET · HACKATHON TRACK
09
TRACK 02 · SINGULARITYNET METTA
MeTTa as a graph
relationship database.
Model entities and the relationships between them, then write rules that traverse or score that graph to produce an answer a human can inspect. Both training exercises are this shape: the connection finder walks relationships, the recommendation engine scores them.
HARD REQUIREMENT
MeTTa holds the relationships and does the reasoning. A model or an API may sit on top for input and phrasing, but the answer comes out of the graph — and you must be able to display the path or the rule that produced it.
A deliberate scoping: MeTTa does considerably more than this. The graph framing is here to give teams a finishable target in the time available.
BUILD YOUR OWN IF IT
It has a genuine graph underneath it.
It needs multi-hop traversal or rule-based constraints, not a single lookup.
It produces an answer whose reasoning can be shown.
BASIX.MARKET · HACKATHON TRACK
10
02
TRACK 02 · EXAMPLE IDEAS
Three things worth building.
Expertise Locator
Find the shortest path from a question to the person who can answer it, through shared departments, cohorts, publications, past projects and topics. The path is the deliverable, not just the name.
Local Trade Connection Finder
Map suppliers, producers and buyers in a local economy and surface viable two- and three-hop trades where no direct match exists.
Constrained Recommender
The movie engine applied to a catalogue with hard rules — compatible parts, dietary restrictions, prerequisite ownership — where constraints beat similarity scoring.
MeTTa = Meta-Type Talk
The language of the Hyperon cognitive stack — how agents represent, reason over and exchange knowledge.
BASIX.MARKET · HACKATHON TRACK
11
TRACK 03 · OMEGA AI AGENTS
Agents that remember,
and show their work.
Build stateful agents on Omega (Omega), a neural-symbolic agent framework on the Hyperon AGI stack, designed for agents with auditable reasoning and long-term persistence. Target users are small businesses, technical teams and communities.
HARD REQUIREMENT
Built with Omega, demonstrating all three of: state that survives across sessions, a decision a human can audit, and memory that changes future behaviour. An agent that forgets between runs has not used the framework.
RESOURCES · BGICOMMONS.ORG/RESOURCES
BUILD YOUR OWN IF IT
It serves a real business, university or community need.
It needs memory between sessions to be useful at all.
It makes decisions someone would want to inspect or dispute.
BASIX.MARKET · HACKATHON TRACK
12
03
TRACK 03 · EXAMPLE IDEAS
Five things worth building.
Lead Qualification Agent
Scores and tracks prospects across multiple separate sessions, remembers prior conversations, and hands off to a human with the full decision trail attached.
Small Business Ops Copilot
Holds supplier terms, reorder points and outstanding invoices in long-term memory, and acts on them without being re-briefed each time.
Technical Risk Review Agent
Audits a stated stack or codebase and produces a prioritised risk register with the justification for each ranking. Narrower and more finishable in a weekend.
Cultural IP Registry Agent
Registers public art, performances or community assets, builds the provenance record over time, and produces proof-of-value and sponsorship documentation.
Sponsorship Activation Agent
Tracks what a sponsor was promised, what was delivered and what evidence exists for each — maintained continuously rather than reconstructed at the end.
BASIX.MARKET · HACKATHON TRACK
13
WHAT STUDENTS EXPERIENCE
They are not attending a hackathon.
They are entering the infrastructure.
The same pipeline we sell to businesses, run end to end on a cohort in forty-eight hours.
01
Assess
02
Onboard
03
Build
04
Submit
05
Review
06
Showcase
07
Continue
EVERY STEP LEAVES AN ARTEFACT — AN ASSESSMENT, A SUBMISSION, A REVIEW, A RECORDING.
BASIX.MARKET · HACKATHON TRACK
14
PRIZE POOL
More than $1,000
across 11 winning teams.
Spread deliberately wide. Eleven teams leave with a result, a recorded showcase and a reason to keep building on the stack.
$150
4 TEAMS
$100
6 TEAMS
$50
6 TEAMS
BASIX.MARKET · HACKATHON TRACK https://basix.market/lms/hackathons
15
THE BUSINESS PROOF
We don't only talk about AI readiness.
This is what a prospect sees when they ask whether any of it is real.
ONBOARD
real students
ACTIVATE
real schools
TEST
real agent workflows
CONVERT
hackathons into pipeline
A hackathon is the cheapest honest demonstration of an operating model. This one runs the model on a live campus, on camera.
BASIX.MARKET · HACKATHON TRACK
16
WHAT THE WEEKEND PRODUCES
Three pipelines out of one weekend.
The hackathon is not the product. It is the intake mechanism for three of them.
Talent
Assessed, onboarded developers with graded work in the marketplace — the bench we staff client engagements from.
Product
Agent workflows stress-tested by builders against briefs written from real business problems.
Business development
Universities activated, partners on camera, and a portfolio of service concepts a prospect can look at.
[ PLACEHOLDER — DROP IN COHORT NUMBERS AND A NAMED STUDENT OUTCOME AFTER SEPTEMBER 30 ]
SJIT CHENNAI · SEPTEMBER 29–30 https://basix.market/lms/hackathons
The AI infrastructure layer
is being built this month.
Sponsor a track, set a brief, recruit from the cohort, or put your business problem in front of eleven teams for a weekend.
BASIX.MARKET
SingularityNet.io
OMNIVERSITY