This year, Iām focusing on improving how I build API automation in Postman ā writing stronger assertions, cleaner test scripts, and using variables and conditional logic more effectively. I also want to make better use of Agent Mode and connect my collections to CI/CD pipelines so tests can run automatically on every build or deployment.
I chose this because APIs are a big part of my day-to-day testing, and good automation helps catch issues early and reduces repetitive manual checks. Running tests in CI/CD also makes failures more visible and gives the team faster feedback.
Unpopular opinion: most developers suck at explaining what they built. Me included.
I can write good code, fix bugs, make things faster - but explaining my work or writing docs that people can actually use? Thatās where I fall short.
Wake-up call happened recently. Spent two weeks building an API integration, wrote docs I thought were clear, and the frontend team still asked me the same questions 15 times. Thatās on me, not them.
This year Iām getting better at:
Writing docs that are actually useful (not just technically correct)
Creating clear examples so people donāt get stuck
Explaining tech stuff in ways everyone can understand
Why this matters: Postmanās collections and auto-docs are super helpful - they handle all the technical heavy lifting and give you solid API references without the manual work. But hereās what I need to add on top: the why behind decisions, the gotchas to watch for, the real examples that help people actually use the API. That extra context is what Iām bad at.
Good code that people canāt figure out how to use is still bad code. The bigger our systems get, the more this communication skill matters.
This year, I should sharpen narrative leadership - the ability to turn complex technical and ecosystem realities into clear, compelling stories that drive alignment and decisions.
Because as a PM, impact no longer comes from knowing more, but from shaping what others believe, debate, and prioritize; APIs and platforms succeed on shared understanding and conviction, and by owning the narrative I can multiply my influence, reduce friction across teams, and make good decisions feel obvious without adding to my workload.
I am already experimenting with AI tools like Claude Code to make my non-coding work dramatically more efficient (writing, analysis, synthesis, planning). Iād like to master that so I can free time and cognitive space to operate at a more strategic level and amplify my overall impact.
Skill Iām committing to this year:
Backend engineering from first principles. why: Iāve noticed that most backend failures Iāve seen arenāt about missing features they come from weak fundamentals. State is implicit, invariant arenāt enforced and once something goes wrong in production, the system becomes hard to fix easily
This year, Iām focusing on backend from first principles. Before scaling or automating anything, I want to be explicit about state, data flow, and failure modes, and design systems that stay understandable when things break
Iām experimenting with languages like Go as constraints to force clearer thinking around state and concurrency
Breaking a Bad Assumption Iāve Had as a Backend Developer
I built APIs expecting people to read docs, understand edge cases, and use them ācorrectly.ā
In reality, many consumers were QA engineers, support teams, test suites and now, increasingly, AI agents.
That changed how I think about my work.
The skill Iām sharpening this year: Designing APIs for humans and machines.
My focus is on:
APIs that explain themselves through concrete examples
Safer defaults and error messages that guide, not blame
Validating usability by testing from a non-developer perspective
As AI starts calling our APIs as much as people do, ambiguity becomes a bug.
Tools like Postman help me see APIs the way others (and AI workflows) will while contracts and mocks make expectations explicit early. Also Since AI is shaping how we all work , i am gonna work on my skills of using AI is the most effective way possible wherever required : )
Happy new year! The skill iām focusing for in 2026 is System Design.
What I am doing: Most developers (including me) rush to start typing code immediately. This year, I am learning to stop and plan first. I am focusing on drawing diagrams, choosing the right databases, and planning how the whole system connects before I write a single line of code.
Why: I realized that coding is actually the easy part(now with the use of AI). The hard part is building complex structures that doesnāt crash when it gets big.
I want to move from being a ācoderā who types syntax, to an āarchitectā who solves problems. If the blueprint is solid, the coding is easy.
This year Iām trying to actually get good at Postman instead of just poking around in it.
Iāve used it casually for a while, but my collections always turn into a mess. My goal is to learn how to organize them properly from the start: version them, keep them small, and not end up with 47 environment variables that only make sense to me. The skill Iām building is sustainable test organization. I want to avoid creating something that becomes a pain to maintain, while still getting value from the automation features. Iām already running into typical problems: weird Git conflicts with the JSON files, tests that only pass on my machine, and realizing some āautomatedā tests are basically just scheduled manual checks.
Lately Iāve been playing with the Postman CLI and trying to figure out when Flows are worth using versus just writing JavaScript tests. The honest answer seems to be āusually just write the code,ā but thereās something nice about visual workflows for complex integration tests. Iām trying to learn where to draw the line so I donāt end up with something fragile that breaks all the time.
Hello everyone, happy 2026 . One skill Iām committed to improving this year is my understanding of REST APIs, so I can better automate and document data workflows that support effective collaboration between teams.
This year, Iām focusing on mastering the intersection of machine learning engineering, API design, and distributed data processing specifically building scalable, production-ready data pipelines that power realātime ML models.
The biggest challenge isnāt just the models themselves but orchestrating the data flows behind them: designing APIs that serve model inference, creating faultātolerant ETL pipelines with PySpark and Hadoop, and ensuring every dataset is versioned and reproducible.
Most teams operate in silos - data scientists, backend engineers, and infra specialists rarely overlap. I want to bridge that gap by designing dataācentric APIs, building scalable PySpark and Hadoop pipelines, integrating CI/CD for model and API testing, and developing resilient systems where data quality, reliability, and performance are tightly connected.
I chose this path because data science and backend engineering are merging. Companies need engineers who understand the whole lifecycle - from model development to deployment and monitoring. Postman has been central to my API learning journey, and combining that skill with distributed computing and ML will help me design intelligent, scalable systems built for the future.
Excited for whatās ahead - 2026 is just getting started!
State Space Models for Sequence Learning is the field I m going all in 2026
for past few months i was learning how Transformers work which help me realise how llm generate output how it generate tokens and then recently I came across a paper
Its about state space model,a potential alternative to transformers and what really intrigued me about State Space Models is how they handle sequence memory. Instead of attending over every past token like Transformers, they update a compact internal state step by step. That state carries forward information from the past, which makes long sequences much more memory-efficient given how expensive context length has become nowadays, especially at scale, this way of modeling sequences feels fundamentally different and worth going deep on
My focus for 2026 is federated learning, which is a technique to train ai without taking user data. Right now, most companies collect our personal data and send it to a central cloud server to train their models, which creates huge privacy risks. Federated learning flips this by sending the model to the userās phone to learn right there, so the personal data never leaves the device. I want to master this because I believe the best technology of the future wonāt just be smart; it will be private, secure, and respectful of the people using it. Btw, happy new year to allll !!!
This year Iām working on designing APIs that make sense to the people using them, not just APIs that work technically. Iāve realized I tend to structure things around how my database looks or how my code is organized, which means other developers have to think like my backend just to make a simple call. Thatās backwards. Now Iām trying to shift my approach and think from the userās perspective first. What do they actually need, what fields are really necessary, how can I make error messages helpful instead of cryptic. Iām also planning to share my designs earlier and get feedback before everythingās built out. The coding part comes naturally to me, but making APIs that feel intuitive and easy to work with is actually harder than I thought. Thatās the skill gap I want to close this year.
This year, Iām learn how to test API and dApp effectively as a writer and also looking to become an Ambassador or influencer for Postman.
The reason is that Iāve encountered how developers push out API and, upon using them, you witness errors in the API or in the application. As for Postman, Postman amaze me always. As someone who uses Postman every day, I see this new update, and I want people to see it.
Thanks to everyone who joined this weekās challenge.
This weekās winner is @sk-fares whose entry stood out for its balanced focus on strengthening API fundamentals while also improving communication skills.
Big thanks to everyone who took part. The submissions covered an impressive range, from API test automation and observability to AI agents, system design and technical writing. Itās awesome to see commitments to both technical depth and professional growth going into the year ahead
. One skill Iām committed to improving this year is my understanding of REST APIs, so I can better automate and document data workflows that support effective collaboration between teams.
Challenge Winner