AI Agent & Data Engineering Intern – Excel Technologies Ltd. (Full‑time)
Posted: Thu Dec 11, 2025 3:23 am
How to Prepare for the AI Agent & Data Engineering Intern Role at Excel Technologies Ltd.
1. Understand the Role Inside‑Out
- Read every line of the job description several times. Identify the core responsibilities (RAG pipelines, LLM integration, FastAPI, multi‑chain processing, chatbot development, SQL agents, web‑scraping, automation scripts).
- Write a short one‑page summary in your own words describing what a typical day might look like. This will help you speak confidently during interviews.
2. Build a Targeted Portfolio
a. Core‑Skill Projects (must‑have)
1. Python + FastAPI – Create a simple RESTful API that receives a text prompt, calls an OpenAI model, and returns the generated answer. Use async endpoints and demonstrate clean modular code (separate router, service, and models files).
2. Vector Search Demo – Choose one vector store (FAISS, ChromaDB, or Pinecone). Index a small document collection (e.g., Wikipedia snippets) and build a retrieval‑augmented generation pipeline that fetches relevant passages before sending them to an LLM.
3. SQL Agent Prototype – Write a script that accepts a natural‑language query, uses an LLM to convert it into a safe SQL statement, executes it against a sample SQLite/PostgreSQL DB, and returns the result. Include basic row‑level security checks.
b. Bonus‑Skill Projects (highly impressive)
1. LangChain Multi‑Chain – Assemble a chain that first retrieves context, then performs reasoning, and finally formats the answer. Publish the chain as a FastAPI endpoint.
2. Custom Chatbot – Build a small UI (Streamlit or Gradio) that talks with a domain‑specific LLM (e.g., a product‑support bot). Integrate role‑based responses and demonstrate how you would enforce access limits.
3. Automation & Orchestration – Use Python’s `asyncio` or `Celery` to schedule periodic web‑scraping jobs, store the data in a vector DB, and trigger a downstream RAG pipeline.
c. Documentation & Version Control
- Host every project on GitHub with a clear README (purpose, tech stack, setup steps, demo screenshots/video).
- Use meaningful commit messages and follow a consistent branch strategy (e.g., `feature/rag-pipeline`).
- Include a `requirements.txt` or `poetry.lock` file for reproducibility.
3. Strengthen Core Technical Knowledge
| Area | What to Study | Suggested Resources |
||||
| Python best practices | Clean code, type hints, modular architecture | “Effective Python” (Brett Slatkin), Real‑Python tutorials |
| Async programming & FastAPI | AsyncIO fundamentals, dependency injection, background tasks | FastAPI official docs, “Async Python” by Matthew Fowler |
| LLM APIs | OpenAI ChatCompletion, Claude, Hugging Face Inference, Ollama | OpenAI API quickstart, Anthropic docs, Hugging Face Transformers tutorials |
| Vector databases | Indexing, similarity search, metadata filtering | FAISS tutorial (official), ChromaDB docs, Pinecone “Getting Started” guide |
| REST API design | HTTP methods, status codes, OpenAPI spec | “RESTful API Design” by Microsoft, FastAPI OpenAPI auto‑generation |
| Prompt engineering | Few‑shot prompting, chain‑of‑thought, retrieval‑augmented prompts | “Prompt Engineering Guide” (OpenAI), LangChain PromptTemplate docs |
| Basic DevOps (bonus) | Dockerfile creation, container run, simple CI with GitHub Actions | Docker official docs, GitHub Actions “Hello World” workflow |
4. Polish the Application Materials
- Resume: 1‑page, reverse‑chronological. Highlight each project with bullet points that start with strong action verbs (e.g., “Designed”, “Implemented”, “Optimized”). Include specific metrics when possible (e.g., “Reduced query latency by 30 % using FAISS IVF index”).
- Cover Letter: Align your story with the company’s vision (“building the future of intelligent systems”). Mention a concrete project from your portfolio that mirrors a key responsibility. Show enthusiasm for learning new LLM technologies.
- Portfolio Link: Add the GitHub URL next to each relevant project in the resume. Ensure the repositories are public and the code runs without hidden secrets.
5. Prepare for the Technical Interview
1. Self‑Assessment: Practice explaining each portfolio project aloud, focusing on the problem, approach, tools, challenges, and results.
2. Live Coding: Expect a Python coding exercise (e.g., write an async endpoint, implement a simple vector search, or convert a NL question to SQL). Use a whiteboard or shared editor to simulate the environment.
3. System Design (Mini): Be ready to sketch a RAG pipeline or a chatbot architecture, describing data flow, components, scaling considerations, and security (row‑level access).
4. Behavioral Questions: Prepare STAR (Situation, Task, Action, Result) stories that demonstrate curiosity, ownership, teamwork, and rapid learning. Example prompts: “Tell us about a time you had to learn a new technology in a week.”
5. Mock Interviews: Pair with a peer or use platforms like Pramp/Interviewing.io to simulate both coding and design rounds.
6. Learn About the Company & Culture
- Visit the company address (House 02, Road 2, Dhanmondi, Dhaka) virtually via Google Maps to familiarize yourself with the commute.
- Research Excel Technologies Ltd.’s recent news, product releases, or open‑source contributions. Mention any alignment in your cover letter or interview.
- Understand the benefits (two‑day weekend, subsidized lunch, company SIM, possible conversion to regular employee) – this shows you have read the posting thoroughly.
7. Timeline (Sample 6‑Week Plan)
| Week | Goal |
|||
| 1 | Deep‑dive into job description, outline portfolio projects, set up development environment. |
| 2‑3 | Complete Core‑Skill Projects (FastAPI LLM service, vector search RAG, SQL agent). Push to GitHub with documentation. |
| 4 | Build at least one Bonus‑Skill Project (LangChain chain or chatbot UI). Polish README and add screenshots. |
| 5 | Refine resume & cover letter, incorporate project metrics, get feedback from a mentor or career coach. |
| 6 | Conduct mock technical and behavioral interviews, finalize application materials, submit before Dec 31, 2025. |
8. Final Checklist Before Submission
- Resume ≤ 1 page, tailored to the internship.
- Cover letter personalized, ≤ 350 words.
- GitHub profile with at least three relevant repositories, each with a functional README and runnable code.
- Clean, professional email address (e.g., firstname.lastname@domain.com).
- Confirm the online application portal is functional; test file uploads if required.
- Keep a copy of all submitted documents for future reference.
By following these steps you will demonstrate both the technical depth and the proactive mindset that Excel Technologies Ltd. seeks in its AI Agent & Data Engineering Interns. Good luck!
1. Understand the Role Inside‑Out
- Read every line of the job description several times. Identify the core responsibilities (RAG pipelines, LLM integration, FastAPI, multi‑chain processing, chatbot development, SQL agents, web‑scraping, automation scripts).
- Write a short one‑page summary in your own words describing what a typical day might look like. This will help you speak confidently during interviews.
2. Build a Targeted Portfolio
a. Core‑Skill Projects (must‑have)
1. Python + FastAPI – Create a simple RESTful API that receives a text prompt, calls an OpenAI model, and returns the generated answer. Use async endpoints and demonstrate clean modular code (separate router, service, and models files).
2. Vector Search Demo – Choose one vector store (FAISS, ChromaDB, or Pinecone). Index a small document collection (e.g., Wikipedia snippets) and build a retrieval‑augmented generation pipeline that fetches relevant passages before sending them to an LLM.
3. SQL Agent Prototype – Write a script that accepts a natural‑language query, uses an LLM to convert it into a safe SQL statement, executes it against a sample SQLite/PostgreSQL DB, and returns the result. Include basic row‑level security checks.
b. Bonus‑Skill Projects (highly impressive)
1. LangChain Multi‑Chain – Assemble a chain that first retrieves context, then performs reasoning, and finally formats the answer. Publish the chain as a FastAPI endpoint.
2. Custom Chatbot – Build a small UI (Streamlit or Gradio) that talks with a domain‑specific LLM (e.g., a product‑support bot). Integrate role‑based responses and demonstrate how you would enforce access limits.
3. Automation & Orchestration – Use Python’s `asyncio` or `Celery` to schedule periodic web‑scraping jobs, store the data in a vector DB, and trigger a downstream RAG pipeline.
c. Documentation & Version Control
- Host every project on GitHub with a clear README (purpose, tech stack, setup steps, demo screenshots/video).
- Use meaningful commit messages and follow a consistent branch strategy (e.g., `feature/rag-pipeline`).
- Include a `requirements.txt` or `poetry.lock` file for reproducibility.
3. Strengthen Core Technical Knowledge
| Area | What to Study | Suggested Resources |
||||
| Python best practices | Clean code, type hints, modular architecture | “Effective Python” (Brett Slatkin), Real‑Python tutorials |
| Async programming & FastAPI | AsyncIO fundamentals, dependency injection, background tasks | FastAPI official docs, “Async Python” by Matthew Fowler |
| LLM APIs | OpenAI ChatCompletion, Claude, Hugging Face Inference, Ollama | OpenAI API quickstart, Anthropic docs, Hugging Face Transformers tutorials |
| Vector databases | Indexing, similarity search, metadata filtering | FAISS tutorial (official), ChromaDB docs, Pinecone “Getting Started” guide |
| REST API design | HTTP methods, status codes, OpenAPI spec | “RESTful API Design” by Microsoft, FastAPI OpenAPI auto‑generation |
| Prompt engineering | Few‑shot prompting, chain‑of‑thought, retrieval‑augmented prompts | “Prompt Engineering Guide” (OpenAI), LangChain PromptTemplate docs |
| Basic DevOps (bonus) | Dockerfile creation, container run, simple CI with GitHub Actions | Docker official docs, GitHub Actions “Hello World” workflow |
4. Polish the Application Materials
- Resume: 1‑page, reverse‑chronological. Highlight each project with bullet points that start with strong action verbs (e.g., “Designed”, “Implemented”, “Optimized”). Include specific metrics when possible (e.g., “Reduced query latency by 30 % using FAISS IVF index”).
- Cover Letter: Align your story with the company’s vision (“building the future of intelligent systems”). Mention a concrete project from your portfolio that mirrors a key responsibility. Show enthusiasm for learning new LLM technologies.
- Portfolio Link: Add the GitHub URL next to each relevant project in the resume. Ensure the repositories are public and the code runs without hidden secrets.
5. Prepare for the Technical Interview
1. Self‑Assessment: Practice explaining each portfolio project aloud, focusing on the problem, approach, tools, challenges, and results.
2. Live Coding: Expect a Python coding exercise (e.g., write an async endpoint, implement a simple vector search, or convert a NL question to SQL). Use a whiteboard or shared editor to simulate the environment.
3. System Design (Mini): Be ready to sketch a RAG pipeline or a chatbot architecture, describing data flow, components, scaling considerations, and security (row‑level access).
4. Behavioral Questions: Prepare STAR (Situation, Task, Action, Result) stories that demonstrate curiosity, ownership, teamwork, and rapid learning. Example prompts: “Tell us about a time you had to learn a new technology in a week.”
5. Mock Interviews: Pair with a peer or use platforms like Pramp/Interviewing.io to simulate both coding and design rounds.
6. Learn About the Company & Culture
- Visit the company address (House 02, Road 2, Dhanmondi, Dhaka) virtually via Google Maps to familiarize yourself with the commute.
- Research Excel Technologies Ltd.’s recent news, product releases, or open‑source contributions. Mention any alignment in your cover letter or interview.
- Understand the benefits (two‑day weekend, subsidized lunch, company SIM, possible conversion to regular employee) – this shows you have read the posting thoroughly.
7. Timeline (Sample 6‑Week Plan)
| Week | Goal |
|||
| 1 | Deep‑dive into job description, outline portfolio projects, set up development environment. |
| 2‑3 | Complete Core‑Skill Projects (FastAPI LLM service, vector search RAG, SQL agent). Push to GitHub with documentation. |
| 4 | Build at least one Bonus‑Skill Project (LangChain chain or chatbot UI). Polish README and add screenshots. |
| 5 | Refine resume & cover letter, incorporate project metrics, get feedback from a mentor or career coach. |
| 6 | Conduct mock technical and behavioral interviews, finalize application materials, submit before Dec 31, 2025. |
8. Final Checklist Before Submission
- Resume ≤ 1 page, tailored to the internship.
- Cover letter personalized, ≤ 350 words.
- GitHub profile with at least three relevant repositories, each with a functional README and runnable code.
- Clean, professional email address (e.g., firstname.lastname@domain.com).
- Confirm the online application portal is functional; test file uploads if required.
- Keep a copy of all submitted documents for future reference.
By following these steps you will demonstrate both the technical depth and the proactive mindset that Excel Technologies Ltd. seeks in its AI Agent & Data Engineering Interns. Good luck!