Building My Way Into a Job: AI as My Career Coach
As we conclude the first quarter of 2026, and set our sights for the late spring and summer, I find myself back on the hunt for a new job. Software engineering has changed so much with widespread adoption of AI and agentic development tools like Claude, Copilot and more. My career interests have also evolved beyond full stack engineering of web applications to the development and orchestration of agentic systems and models.
AI is not new territory for me career-wise. I have built chatbot projects that lived on Cisco Webex, and used Google’s DialogFlow for conversational models. Additionally, I have created AI web app integrations for clients. Lately my focus has been on AI agents as productivity tools for my development process, and building applications that incorporate various large language models to obtain additional information. From the first time I wired up an API call with an Open AI powered endpoint, I knew that this technology would remake the technical landscape completely.
As I’ve been doing more learning and building, I’ve witnessed the growth of agentic development tools from a niche technology, to dominating just about every engineering conversation I’m having these days. When I decided to throw my hat back into the ring for employment, I knew that this time, my search would be different, and so would my process. My goal is to obtain multiple agentic development contracts, or role offers by mid summer.
From previous job searches, I was used to the endless applications, unclear feedback, and my skills evolving faster than my resume. These pain-points indicated that I was not fully leveraging the tools I knew how to build and work with. For this job search, I decided to take a more agentic system oriented approach. My system has 5 roles:
The Scout - analyzes resume + Goals
The Interviewer - creates mock interviews, system design drills, and behavioral simulations
The Builder - suggestions for practice projects to demonstrate my skills
The Teacher - where I store notes from tutorials, and analyze build-a-long videos from Youtube
The Editor - tailors resumes to specific roles, aligns the language with the job description, overall career narratives
To start the scout process, I turned to ChatGPT to ask which roles aligned with my new interests, and how my existing resume could help me get there. What made this step so impressive to me, was that I asked the agent to justify the recommended roles based on my resume, and it came out with clean polished narratives I can use while interviewing. These narratives directly tied my existing skills and experience to the roles I am pursuing, with true and accurate data from my resume.
I also had ChatGPT list out practice projects to help me impress interviewers, and prepare for the questions and decisions I’d be asked to justify. I most enjoy the development part of software engineering, so clarifying design systems and helping me scaffold these projects was very crucial. ChatGPT also generated a mock interview with answers I could build on to make myself a stronger candidate.
For my builder agents, I utilized both Copilot and Claude code for implementation support. After experimenting with these two for months, I finally have a handle on how to get a project running quickly. Additionally, these tools have helped me refine my system design knowledge and think more as an architect, versus an individual contributor.
My teacher agent has been Google’s NotebookLM. Here, I can take my agent output or even a YouTube tutorial link, and ask for an expounded breakdown, along with source code, to better help my understanding of concepts like advanced retrieval augmented generation. Some topics I have been focused on lately are agentic system design, agentic AI development, and software factories and their usages.
For my editor, I turned back to ChatGPT to help with refining my career narratives in relation to my goals. This agents focus was on asking the deeper why behind my goals, what was I hoping to achieve, and how I would get there.
To demonstrate my agentic development skills, I have built a software engineer interview retrieval augmented training agent called StackReady. StackReady allows users to query across a broad spectrum of topics including data types, algorithms, and programming languages.Truthfully, it took several tutorials before I felt comfortable building my own RAG system. I built a naive RAG first, without a frontend, followed by an advanced RAG that used query expansion through multi-query retrieval to improve the accuracy of responses. Next, I went to Claude to help me plan the app infrastructure including my agent, models, ingestion files. I installed many of the packages that I used in my earlier RAG tutorial projects including: fastapi, lang chain, chromadb for storage, and the openAI library. These libraries form the basis of the stack of the project. After getting the app running locally, I had to create a public version to share it with the world. For deployment, I went with Vercel for hosting the react/next frontend, and RailWay for hosting the python backend. After a lot of back and forth getting my API’s working, environment variables set, and 20+ redeployments, a working version of StackReady is available below. There are limits to the knowledge however. I am still exploring ways to flatten my study_guide text document to make it easier to chunk by the models. I also want to expand into other advanced RAG techniques and include more resources for the models to reference when answering queries.
As I continue refining StackReady and expanding my understanding of advanced agentic systems, one thing has become clear: this job search is no longer just about finding my next opportunity. It has become a proving ground for the kind of work I want to do next. Every project I build, every workflow I improve, and every concept I study is helping me move from traditional software engineering into a future centered on intelligent systems that can reason, retrieve knowledge, and assist people in meaningful ways.
The same tools reshaping the industry are the tools helping me reshape my own career. What once felt like a frustrating cycle of applications and waiting has become a process of experimentation, momentum, and creation. StackReady is more than a portfolio project; it represents the direction I’m headed: building practical AI products that solve real problems while continuing to sharpen my skills as an engineer and systems thinker.
As we move into the summer of 2026, my goal remains clear: secure new opportunities in agentic development, whether through contracts, collaborations, or full-time roles. But just as important, I want to keep building in public, keep learning aggressively, and keep sharing what I discover along the way. If there’s one lesson I’ve taken from this season, it’s this: sometimes the fastest way into your next role is to build your way there.
StackReady Live App: https://stack-ready.vercel.app/
StackReady Github: https://github.com/AllisonMH/stack-ready
I can see why encouraging learning through local participation would invite discussion, since different people will bring different experiences to the same question. Taking part in a local activity can give a classroom topic a practical purpose. I would keep that thought nearby when looking for ideas about Jankowice community education. I do not have a settled view yet, but this seems like a good place to slow down and look more closely. How could a school invite families into a project without making participation difficult?
My first thought about writing clear and useful local advertisements is that the everyday experience deserves as much attention as the larger explanation. Specific descriptions can save both the reader and the advertiser a round of basic questions. This is why I would approach local classified advertisements with a specific question rather than a fixed expectation. A short discussion of an everyday example would make this easier for me to connect with, especially as someone still learning about the subject. Which details should always appear in a short local listing?
My first thought about learning languages through everyday situations is that the everyday experience deserves as much attention as the larger explanation. A familiar situation can give new words a purpose beyond memorisation. That is the kind of detail I would keep in mind while exploring practical language learning. I can imagine different answers depending on the situation, so I would want to understand the context before drawing conclusions. How would you practise vocabulary so that it becomes easier to use in conversation?
What interests me about telling meaningful stories with digital media is how a broad idea becomes part of an ordinary day. The order of images and words can change how an audience understands a story. For me, that is a useful way into the broader subject of digital storytelling techniques. I would probably come back to this question after trying to work through one concrete example rather than deciding immediately. How do you decide which details belong in a short digital story?
When it comes to using a tourist guide to plan a flexible trip, I would rather understand one useful detail properly than rush through a long list of ideas. A useful guide should help visitors understand distances and pacing as well as major attractions. That is the kind of detail I would keep in mind while exploring tourist itinerary planning. I would be interested in a simple example, because it is often the ordinary details that make an idea memorable. What would you check before turning a list of places into a day plan?