I was the student
you're building for.

At 18, I dropped out. No roadmap, no advisor — no one to tell me what the path looked like or what my options even were. I navigated it the only way I could: by doing.

Four years later I'd built AI systems serving 60,000+ users under real production pressure. That experience is not a gap on a CV. It's exactly why I get this problem from the inside — and why I want to help build the tool I never had.

0+

active users, one engineer

0yrs

founding engineer scope

2024 — Present

Khadoom.ai

Founding Engineer · Clicks Express

QATAR

Joined as the first engineering hire — the product was just an idea, zero users, no proven market. Built everything from scratch. Took it to 60,000+ active users in two years as the sole engineer, with every architectural decision mine.

AI pipelines, real-time dispatch, fleet automation

Telecoms IVR, automated messaging, ticket routing

Legal compliance + finance audit tooling

Next.js · TypeScript · Python · FastAPI · PostgreSQL · Redis · AWS

Founded — Ongoing

HEVYF

Founder · hevyf.com · Wyoming LLC

SAAS

Noticed a gap — every team I worked with across freelance and full-time work faced the same problem: critical reporting and automation that never got done because no one had the knowledge, the time, or a dedicated team to build it. HEVYF solves that. Paying customers. Full product ownership.

Bulk personalised outreach — 500+ messages/min from a CSV

Connects to Sheets, Stripe, Razorpay, email, WhatsApp

The contracting entity for US clients — no sponsorship needed

Next.js · FastAPI · Python · OpenAI · Supabase · Stripe

I'm not the deepest expert on any single layer. I'm the engineer who figures out how they all work together — fast, under real pressure, with no team to catch my mistakes.

understand → think → move → validate → iterate

That's how Khadoom.ai went from zero to 60,000 users. Give me a problem I haven't seen before. That's when I'm most useful.

Stack Overflow era foundations. AI-native execution. I started in the Flutter community when shipping meant understanding every frame and every state mutation. That discipline is what lets me use Claude as a multiplier instead of a crutch — I know what the generated code is doing, which means I know when it's wrong.

Technical context — relevant to this role

Mobile (origin)

FlutterAndroid

Languages & runtimes

TypeScriptPythonNode.js

Frontend & fullstack

Next.jsReact

Backend & data

FastAPIPostgreSQLRedis

Infrastructure & auth

AzureAWSAuth0

AI & orchestration

LLM pipelinesRAG architecturePrompt engineeringElevenLabs

Why I built this

I learn by building.
So before I applied, I built.

“I prototyped the hard parts of your problem before applying — student profile inference mid-conversation, safety pipeline architecture, voice that feels like a counselor and not a chatbot. Not to show off what I can build. To understand what it actually requires.”

“Your AI Guide is the product
I needed at 18.”

The navigational burden

Thousands of credential programs, financial aid vehicles, and career tools exist across Washington State — but from a first-gen student's perspective, they're invisible. The people who need guidance most carry the heaviest navigation burden.

I've lived this problem

Self-taught, no advisor, no map — I pieced together my path from a dozen scattered sources. That experience is not abstract to me. It shaped every decision in how I built the prototype you're about to see.

Here's the working prototype
Working prototype

AI student advisor — try it now.

Real Claude responses, grounded in Washington State context. Student profile infers mid-conversation. Safety pipeline visualization is illustrative — shows the layer structure, not production code.The guide reads replies aloud. Pick a scenario or type a question.

AI Career Coach — WA State

Claude · RAG-grounded · FERPA compliant

Start a conversation

Pick a scenario above or type a question. The AI Guide will respond in text and speak to you — like a real career coach.

Enter to send · Shift+Enter new line · Guide reads replies aloud

Student Profile· builds live

Profile builds as you share context

Demo sources — WA resources

WA State sources appear
after each AI response

Safety Pipeline — illustrative

Input scan

COPPA age check · PII detection

Intent classification

Crisis detection · Off-topic filter

RAG retrieval

Pulling WA State sources

Claude generation

Grounded response building

Output validation

FERPA compliance · Fact-grounding

Activates on every message

Claude Sonnet · ElevenLabs voice · WA State context · Prompt caching·Personal prototype by Jasim Ameen — not an official CampusEvolve product

GitLab Foundation Grant · Credential Transparency

The credential graph — visualized.

This is what the credential transparency layer could look like for students. The GitLab Foundation grant is building the data layer — here's what the experience layer might feel like. Click any node to see details.

Career Goal
Credential / Program
Job Role
Career GoalsCredentials & ProgramsJob RolesNursingSoftware DevElectricianTeacherCNA CertificateAssociate Degree NursingCoding BootcampAS Computer ScienceElectrical ApprenticeshipBA EducationCertified Nursing Asst.Registered NurseJunior DeveloperSoftware EngineerJourneyman ElectricianK-12 Teacher

Real WA State data · Grounded in WSAC, WorkSource, and community college catalogs · Click any node to explore

What I'd build

Day one through ninety.

The 2026 pilot deadline is real pressure. Here's exactly what I'd focus on to make it.

Week 1

Listen, don't ship

Understand the AI Guide's real failure modes before writing a line of product code.

  • Review real student session logs — where does the AI hedge, hallucinate, or lose the student?
  • Map the current RAG architecture — what data sources are grounded, what's missing?
  • Meet the Yakima and Port Angeles pilot contacts to understand student context first-hand
  • No new features. Understanding only.
30 days

First real deliverables

Ship the pieces that unblock the 2026 pilot deadline and make advisors' lives easier.

  • WhatsApp integration — meet students where they actually are (not a webapp)
  • Advisor dashboard v1 — see student profiles, flag emotional risk signals, track session history
  • Student document reader — accept and parse financial aid letters, transcripts, FAFSA confirmations
  • Source citation hardening — reduce hallucinated resource links to near zero
60–90 days

The architecture that scales

Build the pieces that make this replicable beyond Washington State.

  • Credential graph explorer — the experience layer for the GitLab Foundation data graph (you just saw the prototype above)
  • State-agnostic knowledge base architecture — swap WA data for any state's resources by design
  • Proactive nudges — FAFSA deadline reminders, application window alerts, enrollment confirmation prompts
  • Voice mode polishing — make the ElevenLabs integration feel as warm as a real advisor, not a chatbot

Long-term vision

The architecture built for Washington can serve every education system worldwide.

The same navigational burden that first-gen students face in Yakima exists in Texas, in Brazil, in South Africa. State-agnostic architecture from day one means this isn't a Washington product — it's a template for every education system that has failed to connect learners to opportunity. The Gates Foundation already knows this. The architecture just needs to be built right.

Why I reached out

I'm not casting a wide net.

I specifically want to work on this — the mission is real, the 2026 pilot deadline is real, and the senior full-stack role is still open.

Your AI Guide is the product first-gen students everywhere need. I want to be part of the team that finishes building it.

The honest case

The objections — and what's actually true.

I know the gaps. I'm not going to dodge them.

Location

The role is US-only. How does the location work?

I run ET-aligned afternoons as my standard schedule — it's exactly how I've been working with US clients for two years through HEVYF LLC Wyoming. You invoice a US company. No sponsorship, no immigration paperwork, no HR complexity. Same setup every time. Zero issues.

Experience

You have 4 years. We want 6–10.

Four years as the sole engineer on an AI platform scaled to 60,000+ active users — every architectural decision mine, zero-to-one, no team to catch mistakes. That forces a level of systems thinking that 10 years of maintenance work doesn't. The prototype you just used was built in four days. That's the argument.

Education

No formal CS degree.

I stepped away before I knew this was my path. I taught myself by shipping. The prototype is the degree. Your job posting doesn't mention one either.

In numbers

0yrs

founding engineer scope

0days

to build this prototype end-to-end

$0

ramp-up on Claude / FastAPI / Next.js

LLC

Wyoming — no sponsorship needed

I've already sent an email to info@campusevolve.ai and a direct note to Maria and Toby. If you're reading this — I'd love to find a time to talk through any of this.

One question for you: what's the biggest unsolved engineering challenge on the AI Guide right now?

Live right now

Talk with Jasim

Jasim Ameen

Founding Engineer

HEVYF LLC · Wyoming · ET afternoons daily

“Four years building AI systems under real pressure, with real users. Ask me anything.”

60,000+ active users built
Claude · FastAPI · Next.js
Flutter → Web → AI

Voice via ElevenLabs · Intelligence via Claude