← Back to Blog
communitySep 2025· 15 min read

Building a Developer Fellowship

Lessons from building Dev Weekends, a program that has helped 800+ engineers in Pakistan advance their careers.

When I joined university, I got lucky. I found mentors early who helped me get on track. That alone made a difference.

Then I joined the Microsoft Student Partner program. Back then, Microsoft had a local evangelism team in Pakistan. I learned from them. How to teach. How to run events. How to get people excited about technology and actually help them move forward.

I took that and ran with it. Built the IT Society at university. Delivered 200+ sessions over four years. Learned what actually works when you're trying to help someone level up, and what doesn't.

By the time I graduated, I had a question: what happens to the people who didn't get lucky? The early career engineers who are already working but don't have time to figure everything out on their own? The self-taught developers who don't know what to learn next?

That's where Dev Weekends started.

The idea was simple. Take everything I learned in four years of evangelism and mentorship and apply it to a new format. We started on weekends because that's when working professionals could actually show up.

The Gap We Saw

Universities in Pakistan, and honestly most places, teach syntax. They don't teach systems thinking. They don't teach you how to debug a production issue at 2 AM or how to read a codebase written by someone who left the company three years ago.

We identified three problems that kept showing up:

The Education Gap. Universities teach you to pass exams. Industry needs you to ship products. The distance between "I can write a for-loop" and "I can architect a system that handles 10,000 concurrent users" is vast. Nothing in the curriculum bridges it.

No Roadmap. Brilliant minds stuck without direction. A kid in Lahore with genuine talent doesn't know whether to learn React or Angular, whether to focus on DSA or system design, whether open source matters or is a waste of time. There's no mentor to say "here's what actually matters."

Learning Alone. Self-teaching is romanticized but brutal. Without community, without accountability, without someone to tell you "you're on the right track," most people burn out or drift.

What We Built Over Eight Years

This thing grew into something I didn't expect.

20,000+ community members across 7 countries. Not just passive followers. People actually engaging. The Discord is active almost every day. Conversations happening, people helping each other debug code at midnight, celebrating wins, sharing rejections and figuring out what to do next.

1000+ freelancers in the community sharing work, helping each other find gigs, learning from sessions on freelance engineering. Recruiters actively rely on our community for hiring. When companies need talent, they come to us.

Five different reading sessions where we meet once a week to go through books or series of engineering topics together over Discord. System design, clean architecture, whatever the group wants to dig into. One thing I learned: people don't just need information, they need a group to process it with.

We've run ongoing series on:

  • Automation testing
  • Deep engineering topics
  • Cloud and DevOps
  • AI and ML
  • Freelance engineering

Not one-off sessions but sustained series where people build real understanding over time.

Mind Masters

This one is different from the technical stuff.

We started Mind Masters because we noticed something: you can't build a career if you're burning out. A lot of our members were technically capable but stuck. Procrastinating. Overwhelmed. Lacking direction.

Mind Masters is about understanding your brain. Building discipline. Being more productive. Finding a true purpose in life.

It's an invite-only community for engineers who understand that growth happens beyond the keyboard. We've had sessions on:

  • Psychology of excellence
  • Time and energy mastery
  • Brain health and mental health
  • How to let go
  • The spiritual dimension of work and life

We do Talk of the Week, Talk of the Month, Book of the Week, Book of the Month. Curated content from the world's best thinkers on performance and meaning.

Aside

The best engineers we know have found balance between ambition and contentment. That's what Mind Masters explores. It's not about hustle culture. It's about sustainable excellence.

The Outcomes

Numbers matter because they represent real people whose lives changed:

  • 200+ placements per year through our network
  • 500+ LeetCode Centurions (probably closer to 700+ but we don't always count)
  • 2 GSoC acceptances
  • 30+ actively contributing to open source
  • 20+ engineers placed at Turing alone
  • 67% career success rate

People landing jobs at startups in Europe, at companies they'd only seen in tech news. In Pakistan's ecosystem, open source contributions, GSoC acceptances, competitive programming culture, this stuff doesn't happen. It's not the norm. We're building something that didn't exist before.

Why We Evolved

Everything we did before worked. I want to be clear about that.

The community model, the sessions, the reading groups, the freelancer network, Mind Masters, all of it. It helped a lot of people. We were placing 200+ engineers a year. Recruiters were coming to us for talent. People were building careers.

But we saw room to improve.

The casual drop-in culture meant some people drifted. They'd attend for a few months, learn some things, then disappear. Not because they weren't talented. Because there was no structure pushing them through the hard parts.

We also didn't have enough data. We knew people were getting jobs, but we couldn't always track it systematically. We couldn't replicate success consistently.

And the mentor pipeline needed work. The senior folks who could guide others were busy with their own careers. We needed a better system to bring them back in.

So we didn't kill what was working. We built on top of it.

The Fellowship

In early 2025, we launched The Fellowship.

Same spirit. More structure. Clear tracks, clear milestones, clear accountability.

Aside

The name matters. A fellowship is something you earn your way through. It's not a subscription. Not a course. Closer to an apprenticeship, but with a cohort going through the same thing together.

The Philosophy: Three Pillars

Before the mechanics, let me explain what we believe. The Fellowship isn't just a curriculum. It's built on a philosophy that took us years to articulate.

Pillar 1: Purpose (The Why) Find your "why" before your "what." We've seen too many talented developers burn out because they were chasing jobs, not meaning. We start with spiritual grounding: why are you doing this? What will keep you going when the LeetCode problem feels impossible and the job rejections pile up?

Pillar 2: Psychology (The Mindset) Your mindset is your most important algorithm. We spend significant time on dopamine management, overcoming procrastination, building discipline, and developing what we call "winner mentality." This isn't motivational fluff. It's practical: how do you structure your day? How do you handle rejection? How do you stay consistent when motivation fades?

Pillar 3: Practice (The Skills) World-class engineering through deliberate practice. DSA, system design, tech stacks. But without the first two pillars, the skills don't stick.

We run three types of content, every single day:

  • Morning: Mindset Talks. Productivity hacks, discipline frameworks, the psychology of high performers.
  • Afternoon: Tech Talks. System design deep dives, industry best practices, technical skills.
  • Evening: Spiritual Talks. Purpose, resilience, giving back, building a life beyond code.

How It Works

Three tracks, each 3 months long. You start where you are.

Beginner Track (Bronze) is for people who've never written code or who've dabbled but never built anything real. We strip away the noise. No "learn 17 frameworks in 30 days" nonsense. Programming logic, web basics, data structures. 30 DSA sessions. 12 engineering deep-dives. 12 major projects. By the end, you have a portfolio. Actual things you built. LeetCode 100-200 problems solved.

Intermediate Track (Silver) is where most self-taught developers get stuck. You know syntax. You've followed tutorials. But you don't know how to think about systems. We focus on architecture, how pieces fit together, advanced algorithms, the problem-solving that technical interviews actually test. LeetCode 300-400 problems. Open source contributions.

Advanced Track (Gold) is for experienced developers. Complex systems. AWS, Docker, Kubernetes, CI/CD. Industry collaboration on real problems. And you start mentoring the folks behind you. Teaching cements your understanding. It's also how the whole thing sustains itself. LeetCode 1000+ problems.

Big Word Alert: DSA

DSA stands for Data Structures and Algorithms. The stuff that trips people up in technical interviews. How do you store data efficiently? How do you solve problems that scale? Not glamorous, but it's what separates "can code" from "can get hired."

The 12-Week Sprint

Every fellow operates on the same cycle:

Weeks 1-4: Foundation & Skill Building Master the development environment, core skills, start building first projects. This is where beginners usually stall on their own, spending months picking a text editor. We don't let that happen.

Weeks 5-8: Open Source & Advanced Skills We push people into the wild. Contributing to open source projects. Reading codebases that aren't yours. Applying to programs like Google Summer of Code, LFX Mentorship, Outreachy. Real software is messy and undocumented. This is where you learn that.

Weeks 9-12: Industry Readiness Career-focused. Mock interviews with real engineers (up to 5 per fellow). Portfolio polish. Remote job applications. Resume optimization with ATS-friendly templates. This is where it becomes real.

Choose Your Path

Not everyone wants the same thing. We built three paths within the fellowship:

Path 1: The Competitive Edge For those who want to compete at the highest level. Master algorithms, crack ICPC regionals, compete in Meta Hacker Cup. Deep algorithmic problem solving, Codeforces rating progression, advanced data structures.

Path 2: The Balanced Engineer Excel in both worlds. Build production-grade products while maintaining strong DSA skills. Target remote roles and side projects. Full-stack product development, GSoC participation, remote job interview prep.

Path 3: The Open Source Path Go deep into open source. Crack GSoC, LFX, Outreachy. Build your reputation in the global developer community. GSoC proposal writing mastery, contribution workflow, building maintainer relationships.

The Clan System

When you join the fellowship, you're assigned to a clan. One of 25 focused groups of fellows. Each clan has 2-3 dedicated mentors tracking your progress weekly. This isn't optional. We review your GitHub activity, your LeetCode submissions, your project progress. If you're falling behind, we know. And we reach out.

The Circles Ecosystem

Beyond the main tracks, we've built specialized communities:

  • Open Source Alpha - Beginner open source contributors
  • Open Source Beta - Intermediate contributors on real projects
  • X-Team - Elite performers building production systems
  • ICPC Team - Competitive programming preparation
  • GSoC Prep - Google Summer of Code preparation
  • Interview Prep - Mock interviews and placement prep
  • Remote Job Placement - Landing international remote opportunities
  • 2-Day Startup - Onsite weekend sprints to beat procrastination
  • Co-Working - Virtual sessions for focused productivity
  • Hackathon Circle - Building flawlessly engineered products
  • Weekend Deep-Dives - Topic-focused intensive sessions

The Mentor Cycle

The model works because we closed the loop.

The people who went through Dev Weekends and landed jobs, they're the ones who understand what actually worked. So we brought them back. Engineers at Microsoft, top-rated freelancers on Upwork, founders, senior developers at Calo, Salla, and startups across the world. All volunteering 10-15 hours weekly because someone once invested in them.

Mentees become mentors. Gold helps Silver helps Bronze. Not charity. A cycle. You take, then you give back.

Why It's Free

We had serious conversations about charging. Other bootcamps charge thousands. They have marketing. They promise outcomes.

But the people who need this most can't afford it. The kid in Lahore whose family makes $400 a month. The self-taught developer in a small town who keeps getting rejected because they don't have the "right" background.

That's who we built this for.

The chain continues. Equipment sponsors help us provide laptops and registration fees for students who can't afford them. 100% of donations go directly to student support.

Aside

This isn't sustainable forever at scale. We know that. We're exploring sponsorships, partnerships with companies who want access to our fellows. But we'd rather figure that out slowly than lock people out now.

Where We Are Now

Eight years in. What started with weekend sessions at university is now running almost every day. Reading sessions. Tech series. Mind Masters. Freelancer forums. The Fellowship. Discord always active.

People completing projects. Posting them on LinkedIn. Getting jobs, internships, landing on platforms, working with top companies, getting into GSoC.

The Dev Weekends spirit is still there. The willingness to ask dumb questions. The late-night debugging. The arguments about whether React is still worth learning.

But now there's more structure. A path through the chaos.

We don't just hang out and code anymore. We build engineers.

How to Join

If you're wondering whether you're "ready," you're not. Nobody is. That's the point.

Join our Discord. Complete the kickoff. Get assigned to a clan. Begin.

It's free. It's intense. It works.

Why We Started

Pakistan has talented engineers, but many struggle with the transition from learning to working. The gap isn't just technical—it's about understanding how professional software development actually works. Bootcamps teach syntax; we needed something that taught judgment.

I was also seeing patterns in engineers I worked with. The same mistakes, the same knowledge gaps. I figured if I was already explaining these things one-on-one, I might as well systematize it.

The Three Pillars

We organized the program around three areas that we saw as interconnected:

Purpose - Understanding why you're doing this work. What kind of engineer do you want to be? What problems interest you? Engineers without clarity here often job-hop or burn out.

Psychology - Building consistency and handling the emotional challenges of the profession. Impostor syndrome, debugging frustration, code review anxiety—these are real obstacles that technical training ignores.

Practice - The technical skills, but taught in context. Not "here's how React works" but "here's how you'd build this feature at a company."

What Actually Works

Small cohorts: We cap at 20-25 people per cohort. Larger groups mean less individual attention, and the peer relationships suffer.

Real projects, not tutorials: Participants work on actual codebases with messy history, unclear requirements, and technical debt. This is what professional work looks like.

Code review as teaching: Every participant gets detailed code reviews. This is time-intensive but incredibly effective. People learn more from feedback on their own code than from any lecture.

Alumni mentorship: Graduates who've landed jobs mentor current participants. They remember the struggles and can offer relevant advice. It also creates a virtuous cycle—people who received help want to give it.

What Doesn't Work

Passive content: Recorded lectures have low completion rates. We tried it. People watch at 2x speed and retain nothing.

Unrealistic expectations: Early on, we promised too much. "Land a job in 3 months!" Some people did, most didn't. Now we're honest: this is hard work over 6+ months.

Ignoring the job search: Technical skills don't automatically translate to offers. We added explicit training on portfolios, interviewing, and networking.

The Time Investment

Running Dev Weekends takes 10-15 hours of my week:

  • Weekly sessions: 3-4 hours
  • Code reviews: 4-5 hours
  • One-on-ones: 2-3 hours
  • Admin and planning: 2-3 hours

This is sustainable alongside a full-time job, but barely. As we've grown, I've had to delegate more. Alumni now handle most code reviews and one-on-ones.

Measuring Success

We track a few metrics:

  • Completion rate: ~30% finish the full certification
  • Placement rate: 59% of graduates get jobs within 6 months
  • Salary improvement: Average 40% increase for those switching jobs

But the numbers don't capture everything. Some people gain confidence to start contributing to open source. Others build relationships that lead to opportunities years later.

Lessons for Others

If you're thinking about starting something similar:

Start smaller than you think: Begin with 5 people. Get the format right before scaling.

Charge something: We charge a modest fee. Free programs have high dropout rates because there's no commitment signal. The fee also filters for seriousness.

Build in public: Document what you're doing. It attracts the right people and creates accountability.

Plan for your exit: The program shouldn't depend entirely on you. Build leadership within the community early.

What's Next

We're exploring a few directions:

  • Specialization tracks (infrastructure, frontend, mobile)
  • Partnerships with companies for hiring pipelines
  • Remote-first cohorts for people outside major cities

The goal has always been to build something that outlasts my direct involvement. We're getting closer.

The Reward

The hours are real, but so are the outcomes. Getting a message from someone who just landed their first tech job, or who finally understands a concept they'd been struggling with—that makes it worthwhile.

Community building is slow, unglamorous work. But it compounds. And in an industry where we often measure success in deployments and performance metrics, it's good to work on something more human.