MNJ Premier Advisory

MNJ PREMIER ADVISORY / LEARNING PATHWAY 01

Build the skills.
Apply the intelligence.

A practical AI & analytics pathway for a complete beginner. Learn in sequence, build useful tools, and finish with a portfolio you can explain.

Start week 1 Download the 29-page guide ↓
YOUR NEXT THREE MONTHS
12weeks
10hper week
120htotal practice & learning

Foundations → Skills → Real-world application

No coding experience needed to startPractical projects every weekCareer · Personal · Entrepreneurship

START HERE

A clear path from curiosity
to practical capability.

Follow the weeks in order. Spend roughly three to five hours on training, then use the remaining time to build and review. Each week totals 10 hours and ends with a deliverable and a skills check.

Set up once

A laptop, internet, one AI assistant, a spreadsheet tool, and a folder for your work. Use the course environment for Python. On Windows, use Power BI Desktop; on a Mac, use the Sheets dashboard alternative.

Keep the route affordable

Free materials form the core route. Course access may require sign-in. If DeepLearning.AI requires payment, use the assigned free PY4E alternative. Paid certificates, premium tools, APIs, and cloud labs are optional.

Make time for practice

Suggested schedule: Monday 2h, Tuesday 2h, Thursday 2h, Saturday 3h, Sunday 1h. Use the final hour to test your work, fix errors, and update your learning log.

Build sound judgment

Use public or synthetic data. Check facts, sources, and calculations. Keep a raw copy. Treat AI as a tutor and collaborator, and be able to explain everything you submit.

This pathway builds an applied foundation for supervised professional projects and useful personal or business tools. Advanced machine learning engineering requires further study and experience.

THE CURRICULUM

One week. One useful step.

01 / UNDERSTAND

Weeks 1-4
AI judgment & data literacy

02 / BUILD

Weeks 5-8
SQL, Python & dashboards

03 / APPLY

Weeks 9-12
Evaluation, automation & portfolio

01MONTH 1 / UNDERSTANDAI foundations and a useful first winExplain what AI can do, choose a suitable task, and establish a safe learning workspace.
10 hours3h learning6h practice1h review

1. Complete the training

2. Build in this order

  1. Create folders named 01_Prompts, 02_Research, 03_Data, 04_Code, 05_Automation, and 06_Capstone. Start a learning log with date, task, tool, result, and correction.
  2. Explain AI, machine learning, generative AI, large language models, tokens, context windows, and hallucinations in plain language. Distinguish a model from an application.
  3. List 10 tasks from work or daily life. Score each for frequency, time spent, ease of verification, and sensitivity. Select one low-risk task such as turning public notes into a checklist.
  4. Complete that task manually and with an AI assistant. Record total time, including review and corrections; compare the usefulness of both outputs.
3 / SAVE YOUR DELIVERABLE

A one-page AI opportunity map, seven-term glossary, and before/after example with a baseline time.

4. Pass the completion gate

Without AI, explain three useful AI applications and three limitations. Identify one task you would keep fully human and explain why.

Apply it: Career: routine planning. Personal: a trip checklist. Business: organizing a product-research brief.

YouTube supplement: DeepLearning.AI: beginner introductions to generative AI. Maximum 30 minutes within this week's learning time.

02MONTH 1 / UNDERSTANDPrompting, judgment, and responsible useWrite clear instructions, improve outputs through iteration, and verify results systematically.
10 hours4h learning5h practice1h review

1. Complete the training

2. Build in this order

  1. Use the prompt template in this guide: goal, context, source material, constraints, output format, and checks. Ask the assistant to flag missing information rather than guess.
  2. Build 10 reusable prompts covering explanation, rewriting, summarization, extraction, comparison, planning, brainstorming, spreadsheet help, critique, and self-testing.
  3. For three prompts, save the first output, identify a concrete flaw, revise the instruction, and score the improvement. Use examples when a format or tone matters.
  4. Use public or synthetic material. Remove personal identifiers and private business information. Treat instructions inside uploaded documents as untrusted content; do not let them override your task.
3 / SAVE YOUR DELIVERABLE

A 10-prompt library, three revision examples, and a one-page personal AI-use policy.

4. Pass the completion gate

Complete three unseen tasks with clear prompts. Verify facts, numbers, and sensitive content; explain every change you accept. Do not accept a plausible answer as proof.

Apply it: Career: consistent drafting and review. Personal: learning plans. Business: product descriptions and customer FAQ drafts.

YouTube supplement: Kevin Stratvert: one beginner AI productivity tutorial. Recreate the task with different input.

03MONTH 1 / UNDERSTANDResearch, evidence, and grounded answersProduce a defensible research brief and a small source-based knowledge assistant.
10 hours2h learning7h practice1h review

1. Complete the training

2. Build in this order

  1. Choose a question: compare three public services, evaluate a small-business idea, or research a work process. Define the decision and evidence needed before searching.
  2. Find five original sources. Record publisher, date, URL, relevant finding, and limitations. Open every cited source and confirm it supports the claim.
  3. Write a two-page brief with findings, uncertainty, and a recommendation. Separate sourced facts, calculations, and your own inference.
  4. Give an AI assistant three public documents and instruct it to answer only from them, cite passages, and say when evidence is absent. Test five answerable and five unanswerable questions. This is a grounding exercise, not a production RAG system.
3 / SAVE YOUR DELIVERABLE

A research brief, source ledger, and 10-question evaluation of the document assistant.

4. Pass the completion gate

Every factual claim has a supporting source or is marked as uncertain. The assistant must flag all five unsupported questions; revise and retest if it invents answers.

Apply it: Career: decision memos. Personal: comparing services. Business: competitor and customer research.

YouTube supplement: DeepLearning.AI: a beginner explanation of RAG or AI evaluation. Focus on the concept, not API implementation.

04MONTH 1 / UNDERSTANDSpreadsheet and data literacyClean a small dataset, calculate meaningful metrics, and distinguish evidence from misleading patterns.
10 hours3h learning6h practice1h review

1. Complete the training

2. Build in this order

  1. Create a clearly labeled synthetic dataset of 100 orders: order_id, date, product, channel, units, unit_price, and unit_cost. Add several deliberate duplicates, blanks, and inconsistent labels.
  2. Keep an untouched Raw tab. Make a Clean tab, define one row and each column in a data dictionary, and document how you resolve duplicates and missing values.
  3. Calculate revenue = units x unit_price; product cost = units x unit_cost; gross profit = revenue - product cost; gross margin = gross profit / revenue. Handle zero revenue explicitly. Gross profit excludes other operating costs.
  4. Create a pivot summary and two charts. Compare mean with median, counts with rates, and correlation with causation. Have AI explain formulas, then verify at least five rows by hand.
3 / SAVE YOUR DELIVERABLE

A clean workbook with data dictionary, quality log, metric definitions, and a one-page decision summary.

4. Pass the completion gate

Reconcile raw-to-clean row counts, explain every removed row, and reproduce five formula results manually. State a limitation of using synthetic data.

Apply it: Career: operational reporting. Personal: spending summaries. Business: sales and margin tracking.

YouTube supplement: Alex The Analyst: beginner spreadsheet analysis; StatQuest: mean, median, or correlation. Watch only what supports your project.

05MONTH 2 / BUILDSQL: ask reliable questions of dataQuery and summarize tables, check joins, and independently explain the result.
10 hours4h learning5h practice1h review

1. Complete the training

2. Build in this order

  1. Practice SELECT, WHERE, ORDER BY, JOIN, NULL handling, expressions, SUM, COUNT, GROUP BY, and HAVING. Start with the built-in SQLBolt datasets.
  2. Write eight original questions using the tutorial tables: two filters, two grouped summaries, two joins, one missing-value check, and one comparison using aggregates.
  3. Before a join, identify the key and whether each table has one or many rows per key. Compare row counts before and after so duplicated matches do not inflate totals.
  4. Ask AI to review one incorrect query. Explain the fix in your own words. Keep queries read-only; never run generated changes on a real production database.
3 / SAVE YOUR DELIVERABLE

A .sql or text file with eight commented queries, outputs, row-count checks, and plain-language interpretations.

4. Pass the completion gate

Solve three new read-only queries without copying an answer. Explain how NULLs and a one-to-many join can change a result.

Apply it: Career: pulling reporting data. Personal: understanding structured information. Business: customer/order analysis.

YouTube supplement: Alex The Analyst: beginner SQL joins and GROUP BY. Pause and write the query before seeing the solution.

06MONTH 2 / BUILDPython foundations with AI as a tutorRead and write small programs while understanding the code you run.
10 hours5h learning4h practice1h review

1. Complete the training

2. Build in this order

  1. Use the course coding environment or a Python notebook. Practice numbers, strings, variables, lists, dictionaries, conditions, loops, and functions with small inputs.
  2. Write a function that calculates an order's revenue and gross profit. Add a check for invalid negative units and missing values.
  3. Test at least five cases: normal input, zero units, zero price, invalid input, and an empty collection. Predict results before running the code.
  4. Ask AI for hints, explanations, and debugging help. Inspect code before running it. Do not paste credentials or execute commands you cannot explain.
3 / SAVE YOUR DELIVERABLE

An annotated notebook with a working calculator, five test cases, and a short explanation of each code block.

4. Pass the completion gate

Without AI, explain a loop, a condition, and a function; modify the program for a new input. Demonstrate that invalid input is handled.

Apply it: Career: repetitive calculations. Personal: simple planning calculators. Business: reusable pricing and margin checks.

YouTube supplement: DeepLearning.AI or Alex The Analyst: beginner Python. Re-type a short example and change it.

07MONTH 2 / BUILDFiles, repeatable analysis, and APIsTurn a CSV into a reproducible summary and understand how applications exchange data.
10 hours5.5h learning3.5h practice1h review

1. Complete the training

2. Build in this order

  1. Export the week 4 clean data as CSV. Use Python's csv module, or pandas if covered in your course, to load it and check column names and row counts.
  2. Create revenue and gross-profit totals by product or channel. Save a summary CSV. Explain how a package differs from the standard library.
  3. Understand an API request, response, JSON, authentication, rate limits, and usage cost. Read a sample JSON response from the course; a paid AI API account is not required.
  4. Re-run the notebook from a clean start and compare its totals with the spreadsheet. Record the input file, packages, transformations, and known limitations.
3 / SAVE YOUR DELIVERABLE

A reproducible notebook, input and output CSVs, and a README with instructions and reconciliation checks.

4. Pass the completion gate

Re-running all cells produces the same summary. Spreadsheet and Python totals agree within explained rounding. Explain any missing or invalid records.

Apply it: Career: repeatable reporting. Personal: reusable household summaries. Business: recurring product/channel analysis.

YouTube supplement: Alex The Analyst: CSV analysis in Python. Use the approach, then apply it to your own dataset.

08MONTH 2 / BUILDDashboards and decision storytellingBuild a clear dashboard that answers a real question and supports a specific action.
10 hours3h learning6h practice1h review

1. Complete the training

2. Build in this order

  1. Choose Power BI Desktop on Windows or Google Sheets in a browser. Use the week 4 dataset, or the official Microsoft sample when following its tutorial.
  2. Create a one-page report with three metric cards or summary cells, a trend chart, a category comparison, and a filter. Include units, dates, metric definitions, and source.
  3. Use AI to critique clarity, chart choice, and misleading scales. Verify the critique yourself and make improvements.
  4. Write a five-sentence narrative: decision, main finding, evidence, uncertainty, and next action. Reconcile dashboard totals to the source.
3 / SAVE YOUR DELIVERABLE

A dashboard file or shareable screenshot using synthetic/public data, plus a one-page recommendation.

4. Pass the completion gate

A reviewer can identify the decision and three key metrics in one minute. Every chart total reconciles; the story does not imply causation from a simple association.

Apply it: Career: executive reporting. Personal: goal tracking. Business: sales and marketing performance.

YouTube supplement: Guy in a Cube: beginner report layout; Alex The Analyst: dashboard projects. Do not require paid sharing.

09MONTH 3 / APPLYChoose and evaluate the right AI approachDistinguish automation, generative AI, and predictive ML; test a solution before trusting it.
10 hours2h learning7h practice1h review

1. Complete the training

2. Build in this order

  1. Classify three needs: exact recurring rules -> conventional automation; drafting/extraction -> generative AI; predicting a measurable outcome from historical examples -> predictive ML. Explain when a spreadsheet is enough.
  2. Learn training vs testing, data leakage, overfitting, baselines, and precision/recall conceptually. Use a fresh evaluation set after changing a prompt; do not claim model-building expertise.
  3. Build an AI classifier for synthetic customer requests using three categories and an Unknown option. Create 20 examples, manually label them, and reserve 10 as unseen tests.
  4. Measure correct classifications / total test items. Count false positives and false negatives for an important category; compare with an always-most-common-category baseline. Inspect every error and define an escalation rule.
3 / SAVE YOUR DELIVERABLE

A solution-choice memo and a test sheet containing expected labels, predictions, errors, baseline, and limitations.

4. Pass the completion gate

Explain why 10 test items provide limited evidence. Identify at least two failure modes and specify which outputs require human review.

Apply it: Career: triage and quality assurance. Personal: organizing information. Business: customer-request routing.

YouTube supplement: StatQuest: train/test split, overfitting, precision, and recall. Stay with conceptual explanations.

10MONTH 3 / APPLYBuild a controlled workflow automationConnect simple tools, handle failures, and keep a human approval step before consequential actions.
10 hours4h learning5h practice1h review

1. Complete the training

2. Build in this order

  1. Draw the workflow: a test form submission or new spreadsheet row -> validate fields -> classify/format -> create a draft review row -> human decision. Choose one trigger and one clear output.
  2. Build it with test accounts and synthetic records. Start with rule-based formatting. If a free connection allows an AI step, add it; otherwise paste a reviewed AI result manually and label that step manual.
  3. Test five cases: valid input, missing required field, duplicate submission, invalid category, and connector failure. Use an idempotency key or unique record ID to prevent duplicate processing.
  4. Record run status, time, operations/credits, and errors. Document how to stop the workflow and recover a failed record. Do not enable automatic email sending, publishing, purchasing, or deletion.
3 / SAVE YOUR DELIVERABLE

A working scenario, workflow diagram, five-case test log, and an operating checklist with approval and stop controls.

4. Pass the completion gate

The valid input reaches review; missing, duplicate, and failed cases are flagged or safely stopped. Explain the trigger, data flow, limits, and human approval boundary.

Apply it: Career: intake and report preparation. Personal: task organization. Business: lead intake and draft follow-up preparation.

YouTube supplement: Use Make Academy's included walkthroughs first. Kevin Stratvert may help with the connected spreadsheet or form tool.

11MONTH 3 / APPLYBuild your capstone for a real use caseCombine AI, data, and workflow skills into one coherent solution with measurable value.
10 hours1h learning8h practice1h review

1. Complete the training

2. Build in this order

  1. Choose one track from the capstone options: professional operations, personal productivity, or small-business growth. Define the user, decision, scope, inputs, output, and success metric.
  2. Reuse earlier work: a source-based AI assistant, a spreadsheet/dashboard, and one repeatable processing step. Keep an editable source file and explain any manual steps.
  3. Run the same task manually on five representative examples; record time and quality. Test the proposed solution on those examples plus at least five edge cases.
  4. Calculate net time saved including review and maintenance. Estimate recurring tool costs. Record privacy, reliability, and operational limits; revise the solution based on errors.
3 / SAVE YOUR DELIVERABLE

A working prototype, dashboard or analysis, test log, cost/value estimate, and draft case study.

4. Pass the completion gate

Another person can follow your instructions and reproduce the core output. The solution addresses the original decision and handles its named failure cases.

Apply it: Choose the track closest to the learner's actual work, household needs, or business idea.

YouTube supplement: Use one targeted tutorial only to resolve a specific blocker. Spend most of the week building.

12MONTH 3 / APPLYDemonstrate, package, and apply your skillsPresent evidence of practical competence and plan the next professional or entrepreneurial step.
10 hours0h learning9h practice1h review

1. Complete the training

  • No new course. Complete and demonstrate your portfolio.

2. Build in this order

  1. Finalize a concise case study: problem, baseline, design, data/source choices, tests, observed results, limits, and next steps. Include screenshots and sanitized working files.
  2. Record a five-minute demo or present live. Show a normal input, an edge case, your human review step, and the actual outcome. Distinguish measured results from estimates.
  3. Have a peer score the project with the rubric. Fix weak areas and repeat the affected checks. If working alone, score once without AI and ask someone to reproduce one task when possible.
  4. Prepare a portfolio page or folder and two truthful resume/project bullets. For entrepreneurship, draft a scoped pilot offer with deliverables, timeline, price assumptions, and success criteria. For personal use, write a monthly review routine.
3 / SAVE YOUR DELIVERABLE

A portfolio package, five-minute demo, scored rubric, two project bullets, and a 30-day application plan.

4. Pass the completion gate

Score at least 80/100 and pass all mandatory gates. Explain your solution without an AI script, reproduce one result, and respond to a new test case.

Apply it: Professional: propose a supervised workplace pilot. Personal: run and review the routine. Business: validate a small pilot with a real user before selling a larger promise.

YouTube supplement: Alex The Analyst: portfolio presentation. Use it to improve your explanation, not to copy someone else's project.

WEEKS 11-12 / THE CAPSTONE

Make it relevant to your life.

Choose one track. Combine an AI-assisted output, a reconciled analysis or dashboard, and a repeatable step. Test it, measure its value, and show what still needs human judgment.

01

Professional operations

Build a weekly performance briefing for a fictional team. Combine a source-based process FAQ, a KPI dashboard, and a draft reporting workflow. Test accuracy and review time.

Portfolio evidence: reconciled dashboard, cited briefing, SOP, and workflow test log.

02

Personal productivity

Build a household planning assistant using synthetic expenses and public reference material. Create a spending/goal dashboard and a weekly checklist. Keep sensitive details out of public demos; do not automate financial decisions.

Portfolio evidence: clean data, repeatable summary, verified checklist, and measured time saved.

03

Entrepreneurship

Build a small-business customer-intake and sales-insight prototype. Use synthetic leads/orders, draft FAQ responses from approved source material, and route draft follow-ups for human approval. Validate assumptions with a prospective user.

Portfolio evidence: intake workflow, sales/margin dashboard, evaluation sheet, and a tightly scoped pilot offer.

Measure net value

Net weekly hours saved = runs per week × (manual minutes − assisted minutes including review) ÷ 60 − weekly maintenance hours.

Record one-time setup effort separately. Label estimates and report errors alongside time savings.

COMPLETION STANDARD

Prove it through your work.

Capstone assessment rubric — 100 points
CriterionPointsEvidence
Problem and business/personal value15A specific user, decision, baseline, and measurable success criterion.
Data and evidence quality20Sources are attributable; transformations and totals reconcile; limitations are explicit.
AI instructions and output evaluation20Prompts are reproducible; outputs are checked on held-out examples; errors are analyzed.
Working solution and reliability20The normal flow runs; named edge cases are handled; another person can reproduce it.
Responsible use and operating controls15Public/synthetic data, clear permissions, human review, and stop/recovery instructions.
Communication and portfolio10Clear case study, five-minute demo, honest results, and useful handoff instructions.

80/100 + every required gate

All claims must be supported and calculations must reconcile. The solution must be reproducible, use data responsibly, retain human approval before consequential actions, and be explainable without an AI script. Fix weak areas and retest.

A provider certificate can supplement the portfolio. This independent guide does not issue an accredited qualification or promise employment.

MNJ PREMIER ADVISORY / WORKING TEMPLATES

Download. Complete. Reuse.

Eight branded working templates, each with instructions, response fields, and review prompts. Use the editable Word version for longer answers, or type into the fillable PDF and save a new copy for each project or record.

Download all eight templates ↓ Download the template user guide ↓

The 10-page user guide explains what each template does, when to use it, how to complete every section, and how the records work together. For a full project, begin with 04 Project Brief. Use 03 to evaluate one AI response and 05 to test the complete solution across cases.

TEMPLATE 01

Prompt Design and Iteration

Define the task, write the prompt, and record one controlled revision. Complete one copy per reusable prompt.

TEMPLATE 02

Evidence and Source Record

Record one claim and its supporting source. Duplicate this template for each material claim; open the original source before approving it.

TEMPLATE 03

Prompt Evaluation Record

Evaluate one prompt against a defined task. Retain failed runs and human corrections alongside the approved result.

TEMPLATE 04

AI Project Brief

Agree on a small, useful problem before choosing tools. Complete both pages and obtain the project owner review before a pilot.

TEMPLATE 05

Test and Evaluation Record

Use page one for each test case and page two for the overall evaluation. Include normal cases, edge cases, and failures.

TEMPLATE 06

AI Workflow Operating Checklist

Use one copy per run or reporting cycle. Mark each stage and record exceptions. Stop if required inputs or approvals are missing.

TEMPLATE 07

Project Handoff and Portfolio

Use page one to transfer the working solution and page two to explain the case study. Share only approved, sanitized evidence.

TEMPLATE 08

AI Pilot Offer

Scope a small engagement for a real user. Separate measured evidence from assumptions and secure approval before implementation.

The PDF fields are interactive and can scroll. Save and reopen your copy to confirm your entries; use a form-capable PDF reader. The pack also includes the AI tools matrix.

THE AI TOOLKIT / 23 CURATED TOOLS

The right tool for the task.

Compare practical uses, free access, paid plan examples, and billing conditions. This is a curated starting point, not an exhaustive inventory or a shopping list. Begin with one assistant and add a specialist tool when a project needs it.

Download the tools matrix ↓

Prices checked 18 October 2026. US dollars unless EUR is shown; no currency conversion. Annual rates require a yearly commitment. Taxes, credits, usage, API charges, and promotions can affect the total. Paid examples do not include every plan. Gamma's paid price could not be numerically verified; use its live quote. Prices are a snapshot, not a live feed.

Showing 23 tools

AI tools, functional uses, and pricing snapshot
Tool and linksFunctionPractical usesFree accessPaid exampleBilling and limits
ChatGPT ↗Pricing source ↗General assistantDraft and edit text, explain concepts, analyze files, research, and assist with code.Yes - Free plan with tool and usage limits.Plus: US$20/month.Monthly. API usage is billed separately.
Claude ↗Pricing source ↗General assistantAnalyze documents, draft and critique writing, plan work, and assist with coding.Yes - limited Free plan.Pro: US$20/month; US$200/year.Annual rate is paid upfront. Usage limits apply.
Gemini ↗Pricing source ↗General assistantResearch, write, reason over files, and use AI alongside Google applications.Yes - basic access and lower limits.Google AI Plus: US$9.99/month; Pro: US$19.99/month.Google One bundles; features and limits vary. Avoid paying twice for overlapping bundles.
Microsoft Copilot ↗Pricing source ↗General assistantResearch, draft content, create images, and assist in eligible Microsoft 365 apps.Yes - Copilot chat; app integration depends on plan.Microsoft 365 Personal: US$9.99/month; Premium: US$19.99/month.Paid examples are Microsoft 365 bundles. AI benefits in shared plans may be restricted to the account holder.
Perplexity ↗Pricing source ↗ResearchSearch the web, compare sources, and create cited research answers.Yes - Standard plan with limits.Pro: US$20/month or US$200/year.Annual charge is upfront. Open and check citations; API access is separate.
Gemini Notebook / NotebookLM ↗Pricing source ↗ResearchAsk questions of source collections, organize research, and create source-based study materials.Yes - standard access with lower limits.Expanded access in Google AI Plus: US$9.99/month; Pro: US$19.99/month.Included in Google AI bundles; not a second charge if already subscribed to the eligible bundle.
Julius AI ↗Pricing source ↗Data and analyticsAnalyze spreadsheets and datasets, create charts, and explain analytical results.Yes - a limited free tier.Plus: US$20/month or US$16/month with annual billing.Credit limits and file retention apply. Independently reconcile totals and methods.
Notion AI ↗Pricing source ↗Writing and knowledgeSearch workspace knowledge, draft documents, summarize meetings, and assist with workspace tasks.AI trial on Free/Plus; ongoing core AI on Business/Enterprise.Business: EUR19.50/member/month displayed.Localized EUR quote from provider; confirm US currency and billing term. Custom Agents add US$10/1,000 credits.
Grammarly ↗Pricing source ↗Writing and knowledgeImprove grammar, clarity, tone, and style; generate and revise drafts.Yes - basic writing assistance and limited AI prompts.Pro: US$30/member/month; US$144/member/year.Annual price equals US$12/member/month, paid upfront. Other bundles have different inclusions.
DeepL ↗Pricing source ↗TranslationTranslate text and documents, adapt wording, and support multilingual communication.Yes - limited translation; paid plan trials are separate.Individual: EUR7.49/month, billed annually.Localized EUR quote; verify US checkout. API and Write Pro are separate products or add-ons.
Canva AI ↗Pricing source ↗Design and imagesCreate and edit social graphics, presentations, brand assets, and AI-generated designs.Yes - Canva Free includes a limited AI allowance.Pro: US$144/year per person.Annual total; equivalent to US$12/month, not a monthly contract. AI credits/add-ons and commercial rights vary.
Adobe Firefly ↗Pricing source ↗Design and imagesGenerate and edit images, video, and audio; support creative production.Yes - limited daily generations.Standard: US$9.99/month.Monthly, 2,000 credits. Premium generations consume credits. Promotional discounts excluded.
Midjourney ↗Pricing source ↗Design and imagesGenerate visual concepts, illustrations, and image variants.No standard free web/Discord plan; separate niji app trial may exist.Basic: US$10/month or US$96/year.Annual total equals US$8/month. GPU usage, visibility, and commercial terms depend on plan.
Gamma ↗Pricing source ↗PresentationsTurn prompts and source material into presentations, documents, and simple pages.Yes - initial AI credits; they do not refresh automatically.Paid Plus, Pro, and Ultra plans; current amount requires live quote.Numerical paid price could not be verified from the public page. Check the linked provider pricing before budgeting.
Runway ↗Pricing source ↗VideoGenerate and edit video and images for creative projects and campaigns.Yes - a one-time allocation of 125 credits.Standard: US$15/month or US$12/month with annual billing.Credits limit generations; annual pricing requires the yearly commitment.
Descript ↗Pricing source ↗VideoEdit video and audio through text, transcribe, and improve recordings.Yes - limited recording, export, and AI allowances.Hobbyist: US$24/person/month or US$16/month with annual billing.Per-person pricing; media minutes and AI credit allowances apply.
ElevenLabs ↗Pricing source ↗Audio and voiceGenerate voiceovers, transcribe speech, dub content, and create audio.Yes - 10,000 credits/month; commercial rights differ.Starter: US$6/month; Creator: US$22/month.Regular monthly prices, excluding first-month promotions. Confirm commercial licensing and voice consent.
Otter.ai ↗Pricing source ↗MeetingsTranscribe meetings, create summaries, and search meeting content.Yes - Basic, including 300 monthly transcription minutes.Pro: US$16.99/user/month or US$8.33/month with annual billing.Per-user pricing. Duration/import limits apply; obtain appropriate recording permission.
GitHub Copilot ↗Pricing source ↗Coding and appsSuggest code, explain errors, help with tests, and assist development workflows.Yes - limited Copilot Free for eligible individuals.Pro: US$10/month.AI credit/usage limits and extra usage charges may apply. Review and test generated code.
Replit ↗Pricing source ↗Coding and appsBuild and host prototype apps and websites with AI assistance.Yes - Starter access with limits; confirm current allowances.Core: US$20/month.Provider announcement confirms the regular monthly rate. Agent, hosting, and usage can add costs.
Make ↗Pricing source ↗AutomationConnect apps, transform data, and build visual workflows with optional AI steps.Yes - 1,000 credits/month.Core: US$12/month at 10,000 credits/month.Displayed monthly tier. AI connections and excess usage can add separate provider/credit costs.
Zapier ↗Pricing source ↗AutomationConnect apps and automate multi-step tasks, data handling, and AI-assisted workflows.Yes - 100 tasks/month; two-step Zaps.Professional from US$19.99/month displayed.Confirm the billing term and task tier; the starting rate may require annual billing. Separate AI/API charges can apply.
n8n ↗Pricing source ↗AutomationBuild flexible automation and AI workflows with advanced integrations.Community self-hosting available; cloud offers a trial, not an ongoing free plan.Cloud Starter: EUR20/month, billed annually.EUR price; executions and AI provider usage are separate considerations. Self-hosting needs infrastructure and maintenance.

Free plans can impose quotas, one-time credits, and limits on commercial use. Self-hosted software still needs infrastructure and maintenance. Check the provider's current terms before using confidential data or buying a plan.

THE TRAINING DIRECTORY

Go directly to the source.

Use the weekly assignment to decide how much of each course to complete. Introductory modules, selected lessons, and reference tutorials are labeled. You do not need to complete every provider's full catalog.

View all 14 course and training links
  1. Google Skills

    Introductory course; Google sign-in may be needed. No cloud lab required in this guide.

  2. Google Skills

    Introductory course; Google sign-in may be needed. Skip paid labs or optional subscriptions.

  3. Microsoft Learn

    Free learning module; sign in to record assessment progress. No Copilot purchase needed to read it.

  4. Anthropic / Claude Academy

    Free course; account required. Apply the exercises with any general-purpose AI assistant.

  5. Microsoft Learn

    Free module. Covers prompts, trusted data, retrieval-augmented generation (RAG), and data security.

  6. Google

    Free reference tutorials. A Google account is needed to create practice spreadsheets.

  7. SQLBolt

    Free browser exercises. Complete introduction and lessons 1-12; lessons 13-18 are optional.

  8. DeepLearning.AI

    Account required. Access and certificates may require a subscription; check enrollment terms. Free alternative: PY4E below.

  9. Charles Severance / PY4E

    Free lesson materials and videos. Follow only the assigned topics in weeks 6-7; do not attempt the entire course.

  10. Microsoft Learn

    Free beginner module. Power BI Desktop practice requires Windows; use the Google Sheets alternative on a Mac.

  11. Microsoft Learn

    Free tutorial and sample data. Complete preparation and report creation; publishing and paid sharing are not required.

  12. Google for Developers

    Free conceptual course, about 20 minutes. No model coding is required.

  13. Google for Developers

    Free conceptual course, about 45 minutes. Choose an appropriate problem and success metric.

  14. Make Academy

    Free training with an account; provider estimates 3.5 hours. App execution has separate usage limits; paid AI/API calls are optional.

Links reviewed on 18 October 2026. DeepLearning.AI availability was checked against its official provider listing where automated access was restricted. Enrollment and paid transactions were not tested. Provider access, curricula, and certificate terms can change.

SUPPLEMENTARY LEARNING

YouTube, with a purpose.

Pick one relevant video, pause often, and reproduce the method with different input. Limit supplementary browsing to about 30 minutes per week within your learning time.

WEEKS 1-3, 6-7, 9

DeepLearning.AI

AI foundations, beginner Python, and practitioner discussions. Choose beginner lessons before technical interviews.

WEEKS 2-4, 10-11

Kevin Stratvert

Practical AI tools, spreadsheets, and everyday workflow tutorials. Reproduce one task on your own file.

WEEKS 4-8, 12

Alex The Analyst

Excel, SQL, Python, dashboards, and portfolio walkthroughs. Use a focused beginner video for the current week.

WEEKS 4, 9

StatQuest with Josh Starmer

Statistics, train/test splits, overfitting, and machine learning explained visually. Start with conceptual videos.

WEEKS 8, 11-12

Guy in a Cube

Power BI report design and practical troubleshooting. Skip advanced administration until you need it.

AFTER WEEK 12

Turn a learning project
into a useful pilot.

In your next 30 days, choose a real user, agree on one measurable outcome, run a supervised pilot, and improve the weakest component. Update your portfolio with observed results and choose your next course based on the work you want to do.

Download the complete guide ↓

Questions about practical AI and analytics? Contact MNJ Premier Advisory.