What Happened After AWS Certification: How Many Freelance Projects Came?

I passed the AWS Solutions Architect Associate exam on a Tuesday. On Saturday, I checked Upwork at lunch.

Nothing.

I don't know exactly what I expected. Some version of "certified now, work follows" was apparently in my head, even though I knew, intellectually, that's not how any of this works. The exam score report had come through at 11:47 pm on a Tuesday. I'd posted the badge on LinkedIn. Three people congratulated me. One sent a thumbs up emoji.

The Upwork inbox had not received the news.

It took 19 weeks to get the first paid project. What happened in between is the part most "I got AWS certified" writeups skip, and that gap is where most of it actually happened.

---

The First 11 Weeks: Actual Numbers

One week after passing: my first Upwork message. $9 per hour for basic EC2 monitoring support. I applied. No reply.

Over the next 11 weeks: 47 proposals sent. Six replies received. Two interview calls completed. Zero projects won.

That is a zero percent conversion rate across 11 weeks of consistent effort. I'm writing the number plainly because most versions of this story don't.

The exam fee had been roughly Rs 12,600. Prep course was separate. Three months of focused study time: gone regardless. Sitting with an empty Upwork inbox after all of that produces a specific kind of frustration — not the dramatic kind, just the quiet grinding kind where you keep refreshing and the number stays at zero.

About week eight, I decided the platform was wrong for me. Spent a week on LinkedIn outreach instead. Sent 23 connection requests with thoughtful notes. Two responses. No conversations went anywhere.

The problem was not the platform.

---

What Was Actually Wrong

My proposals read like job application cover letters.

"I am AWS Solutions Architect certified with hands-on knowledge of EC2, S3, RDS, Lambda, and VPC. I have completed multiple training projects and am comfortable working with..."

Accurate. Organised. Completely interchangeable with every other certified candidate in the same proposal queue.

An AWS freelancer I followed in a Discord community — someone who had been on Upwork for a few years and was earning consistently — said something that I didn't want to hear but couldn't argue with: proposals that lead with credentials make the client do all the thinking. The client has to imagine how your credentials apply to their specific problem. Most clients don't bother imagining. They just open the next proposal.

Actually, that's not quite right. It's not just about starting with the problem. It's about already having an opinion about it.

---

What Changed

Around week 12, I rewrote how I wrote proposals. New approach: read the job post carefully, think about what the actual architecture problem probably is, then write as if thinking through it out loud.

Before: "I am AWS certified and can help with your cloud infrastructure needs."

After: "Your description mentions the application going down during traffic spikes. That pattern usually means either fixed-capacity EC2 without auto-scaling, or RDS connection exhaustion under load — two different problems with two different fixes. Happy to ask a couple of questions to confirm which one before we talk about scope."

The second version takes longer to write. But it shows something the first version never could: that I read their post and thought about it specifically.

Alongside this, I built a portfolio. Deployed a 3-tier application on AWS, documented the architecture decisions, wrote a CloudFormation template, pushed everything to GitHub. Wrote a cost optimisation case study for a hypothetical e-commerce workload with before-and-after numbers. Linked both from my Upwork profile.

No client ever asked to look at them. But having them changed how I wrote proposals, because now I had something concrete to think from rather than credentials to list.

---

The First Real Project

Week 19. A US-based startup needed staging and production environments fully separated — AWS account isolation, IAM policies, environment-specific parameter stores. About 13 hours of actual work.

Rate: $16 per hour. Total: roughly Rs 18,700.

Not impressive as a number. That client messaged again 3 weeks later for similar scope, which came to about Rs 24,300. Both came from the proposal approach described above. Not a referral, not a connection, not inbound. A specific proposal that engaged with their actual problem.

---

11 Months In

I'm writing this 11 months after that first certification. A second one came in the middle — Developer Associate. Genuinely, the first few rejection cycles taught me more than the second exam did.

Current AWS freelance income: roughly Rs 41,000 to Rs 53,000 per month, part-time alongside a full-time job. About 15 to 18 hours per week.

9 projects total. 3 repeat clients. Average project around Rs 29,000.

None of those numbers are from the "quit your job and go full-time" category. But the hours-to-income ratio is better than anything else I've found that fits around a day job, and every project has been fully remote.

The thing that nobody in the certification-to-earnings content says out loud: certification is a door, not a career. What you do in the first few weeks you walk through it determines whether the door leads anywhere useful. Mine opened to 11 weeks of zero-conversion proposals before I understood what I was actually doing wrong.

That timeline is probably closer to average than the highlight reels suggest...