A unified 12-week, 72-hour engineering studio in Dhaka — uniting live Silicon Valley instruction, 12 integrated courses (ADS-101 to ADS-112), 5 mastery gates, and an enterprise capstone defense.
LEARN LIVE FROM SILICON VALLEY ENGINEERS
Direct live instruction from senior AI engineers in the San Francisco Bay Area, broadcast into our physical Dhaka training studio with hands-on local mentorship.
Live US Concept Studios
1.5h / WeekInteractive live sessions with Silicon Valley practitioners covering mathematical intuition, neural architectures, RAG grounding, and failure modes.
On-Site Dhaka Lab & Build Mentors
4.0h / WeekDedicated local mentors work beside your desk during Guided Labs (2h) and Applied Builds (2h) for real-time debugging and pair programming.
48-Hour Rubric & GitHub PR Critiques
0.5h ReviewEvery week produces one tangible GitHub artifact graded within 48 hours against our 5-dimension rubric (80%+ mastery threshold).
AI & Data Science Mastery Program
FROM QUANTITATIVE FOUNDATIONS TO PRODUCTION-MINDED AI SYSTEMS
By the end of the program, students will be able to frame data and AI problems, build reliable data pipelines and predictive models, develop generative and agentic applications, and communicate, evaluate, and deploy solutions responsibly in real-world settings.
12 COURSES IN ONE UNIFIED PATHWAY
Filter by mastery level below to explore how ADS-101 through ADS-112 take you from beginner statistics and Python to autonomous agents and enterprise AI.
Mathematics & Statistics for Data Science
Functions, vectors, matrices, probability, distributions, sampling, estimation, hypothesis testing, correlation, and regression intuition.
Python Programming for Data Science
Python syntax, modular functions, data structures, debugging, Jupyter notebooks, NumPy, pandas, visualization, and clean coding.
Intro to Machine Learning & Data Science
End-to-end data workflow: problem framing, structured vs. unstructured data, leakage prevention, baseline modeling, and communication.
Fundamentals of Artificial Intelligence
AI problem-solving, search, planning, constraint satisfaction, graphical models, logic, and Python-based AI reasoning.
Machine Learning & Business Intelligence
Supervised & unsupervised learning, forecasting, classification, clustering, recommendation logic, and executive KPI dashboards.
Data Engineering
ETL/ELT ingestion, SQL warehousing, dimensional schema design, automated data quality checks, lineage, and scalability.
Deep Learning
Neural networks, backpropagation, optimization, regularization, CNNs, sequence models, and representation learning in PyTorch.
Natural Language Processing
Tokenization, dense embeddings, named entity recognition, sentiment, transformer architectures, and slice-based NLP evaluation.
Generative AI-Driven Intelligent Apps
Prompt engineering, hybrid retrieval-augmented generation (RAG), structured JSON outputs, evaluation harnesses, and safety guardrails.
Agentic AI
Goal-driven agents with planning loops, deterministic tool calling, memory stores, orchestration, and human-in-the-loop bounds.
Enterprise AI
Enterprise architecture, legacy integration, AI governance, privacy/security, drift telemetry, cost controls, and ROI modeling.
AI & Data Science Capstone Studio
End-to-end integration of problem framing, data pipeline, model/agent service, risk report, deployment plan, and executive demo.
FOUR PHASES & FIVE COMPETENCY GATES
Every phase pairs conceptual depth with a concrete portfolio artifact, verified across five formal competency gates.
PHASE 01 · WEEKS 1–3
FOUNDATIONS
ADS-101 · ADS-102 · ADS-103Build quantitative, programming, and data reasoning with Python, NumPy, pandas, and statistical inference.
PHASE 02 · WEEKS 4–6
MODELING & DATA SYSTEMS
ADS-104 · ADS-105 · ADS-106Move from classical ML and Business Intelligence into reproducible SQL/ETL pipelines and quality gates.
PHASE 03 · WEEKS 7–9
MODERN AI
ADS-107 · ADS-108 · ADS-109Develop PyTorch deep learning models, NLP transformers, and grounded Generative AI RAG applications.
PHASE 04 · WEEKS 10–12
ENTERPRISE DELIVERY
ADS-110 · ADS-111 · ADS-112Orchestrate tool-using Agentic AI, establish enterprise governance & monitoring, and defend your Capstone.
FIVE PATHWAY MASTERY GATES (GATE 0 TO GATE 4)
Diagnostic + Setup
Baseline quiz, BYOD environment setup, domain problem selection & responsible-AI boundaries.
ADS-101 to ADS-103
Python fluency, statistical reasoning, data cleaning, EDA report & defensible problem statement.
ADS-104 to ADS-106
Evaluated baseline model, documented tradeoffs & reproducible SQL/ETL pipeline with quality tests.
ADS-107 to ADS-110
Deep learning/NLP or grounded GenAI & agentic prototype evaluated for quality, reliability & risk.
ADS-111 to ADS-112
Enterprise governance, monitoring spec, ROI case & live Week 12 Capstone executive defense.
THE 6-HOUR WEEKLY OPERATING RHYTHM
Every week follows a disciplined 6-hour studio cycle — with concept checkpoints every 30–45 minutes and rubric feedback within 48 hours.
CONCEPT STUDIO (25%)
Interactive explanation of mental models, mathematical intuition, architecture patterns, and real failure modes.
Checkpoints Every 30–45mGUIDED LAB (33%)
Step-by-step implementation with starter notebooks, live code walkthroughs, and debugging checkpoints.
Starter Notebooks & TestsAPPLIED BUILD (33%)
Solve realistic dataset scenarios at our Dhaka studio (BYOD laptop + external desk monitor), producing a weekly artifact.
1 Weekly GitHub ArtifactREVIEW & DEMO (9%)
Concept check, rubric review, short peer demo, and preview of how the next week connects.
48h Rubric Feedback LoopWEEK-BY-WEEK 72-HOUR SCHEDULE
Filter by phase or click any week below to inspect integrated courses, topics, 6-hour allocation, and mastery checks.
Topics & Activities: Program onboarding & Gate 0 diagnostic; functions, vectors, matrices, descriptive stats; Python basics, notebook workflow, control structures, and data types.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Probability, distributions, sampling, hypothesis testing, correlation/regression intuition; NumPy, pandas, joins, missing data, and visualization.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Problem framing, structured vs. unstructured data, train/test split, leakage prevention, baseline regression/classification, metric selection, and communicating findings.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Search, reasoning, constraints, AI landscape; supervised learning, decision trees, ensembles, classification/regression, and feature pipelines.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Demo #1 (6h Total)
Topics & Activities: Clustering, segmentation, anomaly/recommendation concepts, forecasting, KPI design, dashboard logic, and translating model outputs into business actions.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: ETL/ELT concepts, SQL/warehouse modeling, schema design, data quality checks, lineage, security, privacy, and scaling analytical pipelines.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Neural network fundamentals, activations, loss functions, optimization, regularization, CNN/RNN intuition, representation learning, and training diagnostics.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Text preprocessing, tokenization, embeddings, NER, sentiment/classification, transformer concepts, and practical NLP evaluation.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Demo #2 (6h Total)
Topics & Activities: LLM capabilities/limits, prompting patterns, retrieval-augmented generation (RAG), structured outputs, evaluation sets, hallucinations, latency/cost, and safety guardrails.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Agent architectures, planning/reasoning loops, tool calling, memory, multi-step workflows, human-in-the-loop controls, agent evaluation, and operational boundaries.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Enterprise architecture, legacy integration, governance, risk/compliance, model monitoring, drift, security, change management, and ROI/value case.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Review (6h Total)
Topics & Activities: Final system integration, evaluation review, documentation polish, executive storytelling, live system demonstration, peer critique, and rubric defense.
6-Hour Weekly Load: 1.5h Concept Studio · 2.0h Guided Lab · 2.0h Applied Build · 0.5h Demo #3 (6h Total)
CAPSTONE DELIVERABLES & MASTERY RUBRIC
Students select a domain problem by Week 3, refine the data/model pipeline by Week 6, add modern AI or agentic capabilities by Week 10, and defend the complete solution in Week 12.
Problem & Value Brief
User/stakeholder need, measurable success metric, operational constraints, and responsible-AI boundaries.
Reproducible Data Pipeline
Documented ingestion, cleaning, transformation, SQL warehouse schema, and automated quality checks.
Model or AI System
Classical ML model, deep learning/NLP component, grounded RAG application, or bounded agent workflow.
Evaluation & Risk Report
Quantitative metrics, slice error analysis, failure modes, bias/privacy/security review, and guardrails.
Enterprise Deployment Plan
Architecture diagram, drift & latency monitoring metrics, fallback behavior, cost notes, and adoption workflow.
Executive & Technical Demo
Clean GitHub repository, concise stakeholder presentation, live system walkthrough, and next-step roadmap.
5-DIMENSION MASTERY RUBRIC
| Dimension | Exceeds Standard | Meets Standard (Required) | Needs Remediation |
|---|---|---|---|
| 1. Technical Implementation | Clean, modular, reproducible code with tests and clear instructions. | Working notebook/app with minor cleanup needed; steps are documented. | Code is incomplete, non-reproducible, or copied without understanding. |
| 2. Data & Evaluation | Strong leakage control, appropriate metrics, slice/error analysis, and limitations. | Correct split and baseline metrics with basic interpretation of errors. | Uses inappropriate metrics, leaks data, or makes unsupported claims. |
| 3. AI Engineering | Thoughtful model/prompt/RAG/agent design with guardrails and cost/latency awareness. | Functional AI component with clear inputs, outputs, and basic safety checks. | Fragile demo with no evaluation, grounding, or operational controls. |
| 4. Enterprise Readiness | Clear governance, security, monitoring, integration, and business value logic. | Identifies main deployment requirements, risks, and stakeholder impact. | Ignores privacy, security, monitoring, or real operational constraints. |
| 5. Communication | Crisp story for both technical and business audiences; defends tradeoffs well. | Explains what was built, how it performed, and what should happen next. | Cannot explain methodology, results, or business relevance clearly. |
WORK FOR USA-BASED COMPANIES VIA TETON
Teton serves as your official US contracting partner. Top graduates are contracted directly through Teton to work remotely on production AI and data engineering projects for USA-based companies.
USA Tech Enterprises
Production AI, RAG & Data RoadmapsTop Program Graduates
100% Remote Execution & Guaranteed PayoutsTeton as Your US Contracting Entity
Teton holds Master Services Agreements (MSAs) and vendor contracts with US companies—contracting you directly and handling all international billing and legal compliance.
- ✓ Contracted directly under Teton's corporate umbrella
- ✓ Zero cross-border tax or foreign entity paperwork
Embedded in USA Engineering Teams
Collaborate directly with American tech leads in daily standups and ship production LLM pipelines, agent workflows, and data warehouses for US software products.
- ✓ Direct collaboration with US engineering leads & PMs
- ✓ Production commits on modern Silicon Valley stacks
Guaranteed Payroll & Senior Backing
Receive reliable monthly compensation via local bank deposit, backed by ongoing architectural reviews and mentorship from Teton's senior engineering leadership.
- ✓ On-time monthly payouts with zero banking friction
- ✓ Senior code review & architectural support
WHY STUDY AT TETON?
Teton AI Training Centre is built on the engineering lineage of Teton Electronics. We teach AI as an operational engineering discipline where mathematical rigor, reproducible data pipelines, model latency, and governance matter.
Live US Instruction & On-Site Dhaka Mentors
Live Concept Studios with Silicon Valley instructors broadcast into our physical Dhaka studio with 1:15 in-person lab mentoring.
In-Person Dhaka Studio (BYOD) & External Monitors
Mandatory physical attendance in Dhaka. Bring your own laptop — Teton equips each desk with dedicated external monitors, enterprise Wi-Fi, and power backup.
5 Competency Gates & Executive Capstone Defense
Progress through 5 formal mastery gates with milestone demos at Weeks 4, 8, and 12 and US contractor evaluation.
HOW TO JOIN BATCH 01
Submit Online Application
Share your background and goals for the 12-week AI & Data Science Mastery Program.
Gate 0 Diagnostic Screening
Complete a baseline check on quantitative intuition, Python readiness, and problem framing.
Counselling & Seat Offer
Meet with our academic team at the Dhaka campus to confirm fit and receive your seat offer.
Enrolment & BYOD Setup
Complete tuition payment, configure your laptop environment, and begin Week 1 Foundations.