Traditional analyst workflow
“I produced the analysis.”
1Receive CSV, spreadsheet or database export
2Analyze in R, Excel, SQL or a notebook
3Create charts / dashboard / slides
4Hand results to someone else to operationalize
Economics × Data × Engineering × AI
Grace, you are not learning random tech.
You are learning how to take the quantitative thinking you already have and turn it into repeatable, automated, usable systems that other people can actually work with.
66%
of 2025 U.S. data-scientist postings referenced Python
O*NET / Lightcast
51%
referenced SQL; R remained strong at 34%
O*NET / Lightcast
+34%
projected U.S. data-scientist employment growth, 2024–34
U.S. BLS
84%
of developers use or plan to use AI tools in development
Stack Overflow 2025
The transformation
We are adding leverage, not replacing your identity.
Your economics, statistics, forecasting, quantitative reasoning, business context and R experience are the hard-to-fake part. The new skills let those strengths escape the notebook, spreadsheet or slide deck and become systems.
What you already bring
EconomicsStatisticsForecastingMathematicsRBusiness reasoningInterpretation
engineering + AI literacy→
What becomes possible
Automated analysisAPI-connected dataInternal appsProduction workflowsAI-enabled toolsEnd-to-end ownership
Click the curriculum
Why am I learning this?
Each layer solves a different career limitation. Choose a step to see why it belongs, how deeply you actually need to learn it, and what you should be able to build with it.
Professional foundation
Git + GitHub
Career leverage
What “good enough” looks like
How deeply should you learn it?
Portfolio proof
What you do NOT need
Why this order?
Learn the environment before depending on the language.
GitHub first teaches where professional code lives and how work changes safely. A short web-fundamentals block makes software visible. Python and SQL then become engines inside a system rather than isolated notebook syntax.
01GitHubWork professionally
02HTML / WebSee the interface
03Python + SQLCompute + query
04APIsConnect systems
05AutomationRepeat reliably
06AI / LLMsAugment reasoning
07Apps / AgentsShip capability
One important adjustment: HTML should be a focused foundation, not a long detour before Python. And SQL belongs early alongside Python. The goal is interface literacy, not months of frontend specialization.
Career differentiation
Same analytical brain. Much larger ownership boundary.
The difference is not whether you can calculate a model. It is how far you can carry the work after the model is correct.
Engineering-enabled analyst workflow
“I turned the analysis into capability.”
1Pull data automatically from databases and APIs
2Analyze, model and validate with R, Python and SQL
3Put code under version control with tests and documentation
4Expose the result through an app, service or automated workflow
5Add AI where language reasoning actually improves the system
6Measure whether the system is producing the intended business outcome
The AI-era question
If AI writes code, why learn programming?
Because AI lowers the cost of producing code, but it does not eliminate the need to decide what should be built, inspect whether it is correct, connect it to real systems, understand the data, debug failures, and own the result.
84%
of Stack Overflow’s 2025 respondents said they were using or planning to use AI tools in development. AI-assisted work is becoming normal, not exceptional.
46%
said they distrust AI-output accuracy, versus 33% who trust it. That makes validation and technical judgment more valuable, not less.
66%
named “almost right, but not quite” AI solutions as a major frustration. Knowing enough to diagnose the last 10% is a career skill.
Depth matters
You do not need to master everything equally.
The curriculum should make you dangerous in the skills closest to data and reasoning, competent in the engineering layers, and selectively deep in AI systems that support real projects.
Skill
Learn deeply
Learn enough to build
Do not over-focus on
Economics / stats
✓ Core differentiator
—
Do not abandon it for generic coding
R
✓ Keep it
Integrate with broader workflows
Language tribalism
Python
✓ Deep
Data, automation, backend, AI
Memorizing syntax AI can supply
SQL
✓ Deep
Joins, CTEs, windows, modeling
Vendor-specific trivia
Git / GitHub
Professional fluency
✓ Branch, commit, PR, review
Advanced Git wizardry early on
HTML / CSS / JS
No
✓ Interface literacy
Becoming a frontend specialist
APIs
Strong concepts
✓ Build integrations
Protocol trivia without projects
AI / LLMs
Evaluation + architecture
✓ Build responsibly
Prompt tricks with no system design
RAG / agents / MCP
Later, project-driven
✓ Practical competency
Chasing every framework
The Mac question
Does this curriculum require a monster computer?
No. Hardware should follow the workload. A stronger machine buys headroom for containers, local databases, multiple development services, larger datasets and local AI—but ordinary analysis does not justify extreme specifications by itself.
This is a workload guide, not a recommendation to spend more money.
Balanced / long-term
Strong headroom without workstation excess
6–12 month outcome
One evolving project beats disconnected tutorials.
The best proof that the strategy works is a portfolio where the same economics/business problem becomes progressively more capable as you learn each layer.
Months 1–2
Reproducible analysis
GitHub, README, R/Python, SQL fundamentals. Pick an economic or business dataset and make the project easy for another person to clone and understand.
Proof: public repo + documented analysis
Months 3–4
Connected analysis
Replace manual files with an economic, government or business API. Clean, validate and refresh the data programmatically.
Proof: automated API data pipeline
Months 5–7
Usable application
Put the analysis behind a simple web interface or internal tool. Add testing, configuration and a repeatable run/deploy process.
Proof: working analytical application
Months 8–12
AI-enabled capability
Add an LLM only where language reasoning helps: explain results, search a knowledge base, call analytical tools or assist users. Evaluate its accuracy.
Proof: AI + deterministic analytics in one product
Evidence to practice
Useful starting points
These are not the curriculum by themselves. They are credible places to practice the exact skills above.
GitHub Skills
Hands-on exercises for repositories, branches, pull requests, Markdown, Pages and Actions.
Open GitHub Skills →GitHub Pages
A simple way to publish an early HTML/data portfolio and make the web layer tangible.
Open GitHub Pages →FRED API
Economic time-series data that can turn Grace's forecasting background into an API-connected portfolio project.
Open FRED API docs →U.S. Census APIs
Government datasets suitable for Python/SQL/API projects tied to economics and demographic analysis.
Open Census APIs →The teaching philosophy
“I am not teaching you HTML because I want you to become a web developer. I am teaching you enough of the layers around analysis so that your ideas do not stop at the analysis.”
The target is a person who can move from business question → data → analysis → version-controlled code → automated workflow → usable application → AI-enabled capability → decision. Your economics background remains the brain of the system. The new skills give it reach.
Research basis: U.S. Bureau of Labor Statistics 2024–34 employment projections; O*NET employer-based technology data derived from 2025 Lightcast U.S. job postings; Stack Overflow 2025 Developer Survey; GitHub Octoverse 2025. Percentages describe their respective datasets and should not be read as universal requirements for every job.