Skip to main content
IT Business Analyst at Rocket Global · Hanoi

From a three-line request to a signed-off system.

CRM/ERP business analysis, direct work with US clients, handover to Dev and testing until the client signs off.

CRM & ERP Workflows

Sales processes, EV charging operations, E-learning & Odoo Fit-Gap analysis.

US client point of contact

Sole point of contact, 100% in English: 41 meetings into 128 traced change requests.

Specification & handover

19 modules documented, 234 business rules, Given/When/Then user stories.

Testing & AI

QA/QC Lead: test and CI audits, 901 automated test cases in two weeks.

Figures from real work

Counted from my own documentation repository, Jira and kit repository (06–09/2026); client names withheld.

About me

A Business Analyst with a data and AI background.

Đỗ Hoàng Anh

Đỗ Hoàng Anh

IT Business Analyst · Rocket Global

Vị tríIT Business Analyst
Doanh nghiệpRocket Global
Địa bànHà Nội, Việt Nam
Đào tạoCử nhân AI · ĐH Thăng Long (GPA 8.0)

I am Đỗ Hoàng Anh, an IT Business Analyst at Rocket Global. My core focus is transforming ambiguous business needs into clear specifications, Jira tickets, and testable acceptance criteria so engineering and QA can deliver accurately from sprint one, staying aligned until client signoff.

Trọng tâm nghiệp vụ & phương pháp luận

CRM & ERP Analysis

Designing As-Is/To-Be workflows, business rules, ERD schemas, and wireframes for complex modules.

Tech & AI Agents

Navigating databases, codebases, and harnessing agent skills (Claude Code, Codex) to eliminate 70% of routine tasks.

Global Delivery

Communicating 100% in English with US clients; Chinese at HSK 3.

Trọng tâm hiện tại
Current role
IT Business Analyst · Rocket Global
Working on
US client project · BA Kit & Tester Kit
Daily tools
Jira · Figma · SQL · Claude Code

Six steps, each leaving a document that can be checked

A requirement moves on only when the previous step has its documents and conditions in place. That is why Dev does not have to come back with questions.

  1. 01

    Intake & clarification

    Meet, ask follow-up questions from a checklist, and have the client confirm my understanding in writing.

  2. 02

    Analysis & design

    As-Is/To-Be, impact assessment, business rules, flows and interfaces.

  3. 03

    Handover to Dev

    Split into Epics, Stories and Tasks; testable acceptance criteria.

  4. 04

    Testing & verification

    Agents run what can be automated first; a person verifies and decides.

  5. 05

    UAT & acceptance

    The client runs through a checklist, feedback becomes tickets, the client signs off.

  6. 06

    Measurement & improvement

    KPI dashboard, weekly report, finding repeated steps to standardise.

See the full process and principles

Real work, told decision by decision

Each case covers the context, my role, each step taken and the reasons behind the main decisions.

Documents from this case: User story, Clarifying questionsBusiness Analysis · Delivery100%communication in English

From a US client's request to UAT

Sole point of contact for a US SaaS client, working 100% in English: 41 meetings turned into 128 change requests, 624 Jira tickets and documentation for 19 modules, then testing and UAT.

  • US client
  • English
Documents from this case: Test suite audit, CI auditQA · Quality35 → 276backend tests passing

A green test suite that tested nothing

I audited every test and pipeline: removed 146 fake tests, fixed the root cause that took the backend suite from 35 to 276 passing, ran 901 automated test cases in two weeks and found 10 data-isolation defects before release.

  • QA
  • CI/CD
  • Test automation
146 → 0 fake UI tests removed
Documents from this case: BA Kit, AI draftProcess improvement · AI30self-written skills in BA Kit

BA Kit: from manual BA work to BAs directing AI

The BA team was the bottleneck once Dev had adopted AI. I built a project knowledge vault and 30 agent skills, trained the team and set the principle: AI does the work, a person is accountable.

  • AI Agent
  • MCP
  • Claude Code
−70% time on repetitive tasks
Documents from this case: Business rules, Data modelBusiness Analysis · CRM/ERPCRM + ERPin one system

E-learning system integrated with CRM/ERP

Analysis and design of a system to manage customers, tuition, receivables, classes and teachers; wrote the BRD, PRD, SRS and data model for the Dev team.

  • CRM
  • ERP
  • LMS
BRD · PRD · SRS document set for Dev

All case studies

Data and AI projects

Three university projects in credit risk modelling and image recognition: the groundwork that lets me read data when I analyse a business.

Machine LearningAUC 0.69 · Accuracy 0.75

Lending Club

Credit risk modelling and portfolio insights

A Data Mining team project on more than 2 million loans. I led EDA and feature selection, built an interpretable Logistic Regression model and tuned the threshold to reduce “Fully Paid” misclassifications.

  • Credit risk
  • Logistic Regression
  • Threshold tuning
Data AnalyticsAUC 0.703 · Precision 0.88

Home Credit Default Risk

Default prediction and borrower risk assessment

Gradient Boosting on more than 300,000 records, focused on borrowers aged 28–65 and cash loans. Preprocessing, feature engineering and a 0.162 threshold to screen high-risk applicants.

  • GBM
  • Feature engineering
  • Python
Deep LearningTest accuracy 0.93

Flower Classification

Image recognition with CNN and VGG16

A custom CNN compared with transfer-learning VGG16 to classify five flower species, using data augmentation, batch normalisation and dropout. The final model was packaged as an upload-and-predict app.

  • CNN
  • VGG16
  • Transfer learning

Experience and education

From studying AI and teaching programming to business analysis for international clients.

  1. Rocket Global

    IT Business Analyst

    • 09/2025 – present
    • Hung Yen
    • Turned 41 client meetings into 128 traced change requests; created 624 Jira tickets (47% of the board), working 100% in English.
    + 3 chi tiết công việc
    • Documented 19 modules: 234 business rules, 90 user stories, 729 diagrams; sprint reports and a client bug register built from Jira.
    • Audited tests and CI: backend tests from 35 to 276 passing; 901 automated test cases in two weeks; 10 data-isolation defects found before release.
    • Built BA Kit and Tester Kit (40 self-written skills) and trained 3 BAs and 4 Testers; about 70% less time on repetitive tasks.

    ⭐Sole point of contact for a US client on a 15-repository SaaS platform; QA/QC Lead since 09/2026.

    • Jira
    • Figma
    • SQL
    • Playwright
    • Claude Code
    • MCP
  2. Thinh Cuong Corporation

    Business Analyst Intern

    • 06/2025 – 09/2025
    • Hanoi
    • Mapped As-Is/To-Be processes for electric car and truck sales and charging stations.
    + 2 chi tiết công việc
    • Analysed Odoo CRM/ERP modules, workflows and business rules.
    • Took requirements from 5 departments and wrote handover documents for Dev.

    ⭐Business analysis for electric vehicle sales and charging stations for Thinh Cuong Corporation; fit-gap against Odoo.

    • Odoo
    • CRM
    • ERP
    • HRM
  3. TEKY Academy

    Web Programming Instructor

    • 01/2024 – 10/2025
    • Hanoi
    • Taught web programming fundamentals with Python, HTML, CSS, JavaScript and WordPress.
    + 2 chi tiết công việc
    • Designed hands-on lessons in which students built simple websites and web apps.
    • Guided students through debugging, code review and basic development practices.

    ⭐Helped students go from zero to building their first websites and web apps.

    • Python
    • HTML
    • CSS
    • JavaScript
    • WordPress
  4. A.N Lab

    AI Engineer Intern

    • 05/2023 – 01/2024
    • Hanoi
    • Supported computer vision and NLP models for internal products and client projects.
    + 2 chi tiết công việc
    • Data preprocessing, feature engineering and baseline models for classification tasks.
    • Evaluated model performance and wrote experiment reports for the team.

    ⭐Built and evaluated AI models for real product use cases.

    • Python
    • OpenCV
    • TensorFlow
    • PyTorch
    • Scikit-learn
  5. Thang Long University

    Bachelor of Artificial Intelligence

    • 09/2022 – 2026
    • Hanoi
    • Core courses: Database, Data Mining, Machine Learning, Deep Learning, Big Data.
    + 2 chi tiết công việc
    • Applied ML to real data in projects on credit risk and image classification.
    • Worked in teams to design, build and present end-to-end AI/ML solutions.

    ⭐GPA 8.0/10 · Top 100 outstanding students 2022–2023.

    • Python
    • Pandas
    • Scikit-learn
    • TensorFlow
    • SQL

Full details in the CV:CV in English (PDF)

Professional skills

Business analysis is the core; data and AI are the edge that comes with it.

01

Analysis & documentation

Requirements elicitationAs-Is / To-Be, fit-gapImpact analysisBusiness rulesUser stories, acceptance criteriaBRD, PRD, SRSBPMN, ERDWireframesPermission and traceability matrices
02

Systems & domains

Odoo CRM/ERPCRM, sales processesHRMLMS, e-learningTuition, receivablesElectric vehicles, charging stations
03

Delivery & testing

Agile / ScrumBacklog, sprint planningJira, Plane, TrelloTest case designAPI testingBug reports with evidenceUAT
04

Technical & AI

REST API, webhooksSQL, data modellingReading codebasesClaude Code, Codex, MCPAgent skillsProject knowledge vault
05

Data & ML

PythonSQLPandas, NumPyScikit-learnXGBoost, LightGBMTensorFlow, KerasStreamlit, Gradio

See the documents I write, not just a description of them

Four sample documents that keep the structure and language used in real work; the data has been simulated.

Weekly status report

Weekly status report for the client

Sprint progress report sent to a US client: summary, KPIs, work delivered, blockers and decisions the client needs to make.

Document in English
Meeting minutes · Requirement · Ticket

Meeting minutes to Jira ticket

One meeting passes through three documents: minutes with decisions and open questions, a confirmed requirement, and a ticket that meets the Definition of Ready.

Document in Vietnamese
Test execution report

Agent-assisted test execution report

Results of one test cycle: agents run first and give preliminary results, the tester verifies them, changes results where needed and makes the release decision.

Document in Vietnamese
Business Requirement Document

BRD excerpt

Six representative sections of a BRD for a training centre management system: goals, scope, To-Be process, business rules, permissions, traceability.

Document in Vietnamese

All sample documents

Professional Certifications & Language Proficiency

Professional Certifications

Rigorous foundation in data analytics and machine learning from IBM and Stanford Online.

Credentials & Languages
IBM
IBM & CourseraData Analyst & Visualization
Coursera
IBM & CourseraProfessional Certificate

IBM Data Analyst Professional Certificate

Advanced SQL queries, Python for data analysis, Excel financial modeling, and dashboard design.

Advanced SQLPython for DataExcel DashboardsData Modeling
Xem chứng chỉ gốc
SU
Stanford & VietAIMachine Learning Specialization
VietAI · Stanford
Stanford Online & DeepLearning.AISpecialization

Machine Learning Specialization

Supervised & Unsupervised Learning, Regression, Decision Trees, algorithms, and model evaluation with Andrew Ng.

Supervised LearningUnsupervised LearningModel EvaluationAlgorithms
Xem chứng chỉ gốc
DL
DeepLearning.AI & VietAIDeep Learning Specialization
VietAI · DeepLearning.AI
DeepLearning.AISpecialization

Deep Learning Specialization

Deep neural networks, CNN for computer vision, Sequence Models, and foundational Transformer architectures.

Neural NetworksCNN ArchitectureSequence ModelsPyTorch / TF
Xem chứng chỉ gốc

Language Proficiency

100% English with US clients; Chinese at HSK 3.

Năng lực thực tế
EN
English100% US Client
IELTS 6.0
IELTS 6.0 Academic · Direct US Client Collaboration

Meetings, requirement elicitation, specifications and 624 Jira tickets, 100% in English.

100% US Client DirectTechnical PRDScrum FacilitationCEFR B2
中文
Chinese (中文)HSK 3
HSK 3
HSK 3 · Basic communication

HSK 3 certificate: communicating and reading Chinese in familiar, everyday situations.

HSK 3Basic conversationBasic reading
VI
VietnameseNative
Bản ngữ
Native

Clear logical thinking, precise requirement specification, and rigorous documentation across stakeholders.

NativeRequirements SpecStandardized Docs

Talk about a project or a role

Send a message or get in touch directly through the channels below.

The information you send is used only to reply to this message.