I build governed AI systems people can trust, supervise, and adopt.
I build the human-oversight layer behind agentic workflows — approval gates, context risk, recoverable consequences — and design the provenance and evaluation systems around it. Learning design isn’t my ceiling; it’s my adoption edge.
Governed AI and knowledge systems — plus the learning layer that gets them adopted.
Shipped builds spanning agent supervision, technical upskilling, and high-stakes simulation. Every card opens a full case study, and every case study opens a live system.
AI Workflow Quest
A branching simulation that trains employees to supervise AI agents — approval gates, context risk, and rewindable consequences at the exact decision points a human must keep.
Data Center Technician Upskilling
A three-program upskilling suite that takes a new hire from the floor to a trusted 24×7 shift — with a fully-built flagship (Rack to Runtime) and an interactive 3D 42U rack, grounded in EIA-310-D, ASHRAE, and Uptime/TIA-942 standards.
SiteReady — Scissor Lift Safety
A built, deployable awareness micro-module and printable pre-use job aid that help operators recognize scissor-lift hazards and choose Stop, Fix, or Proceed.
Gas Turbine — 3D Interactive Tutorial
A browser-based 360° cutaway that teaches the anatomy and gas path of a heavy-duty gas turbine — rotate it, strip the casing, and click any component to learn what it does and why it matters.
AeroMind — Drone Flight Trainer
A 3D drone simulator that teaches pitch, roll, yaw, throttle, telemetry literacy, and failsafe behavior — every stick input becomes a visible, explained physics lesson.
Ohio DriveSmart Academy
An interactive React driver-ed app: eight lesson modules, hazard-decision practice, illustrated flashcards, and a guardrailed AI study coach. Educational concept.
From problem to governed, adopted system.
One working loop across AI and learning builds — every stage traces to a risk, an objective, or a measurable behavior. ADDIE and SAM still power the learning layer; the loop around them is wider.
Frame
Problem, risk surface, audience, baseline metric.
Model
Knowledge modeling — facts locked in cited spec libraries, SME sign-off as a gate.
Orchestrate
Workflows, branching, agents, simulations, storyboards.
Evaluate
Scored scenarios, xAPI instrumentation, Kirkpatrick L1–4.
Govern
Human-oversight points, honest scope claims, accessibility, provenance of every figure.
Adopt
Pilot, LMS / SCORM deploy, champions, change comms.
Adult Learning
Relevance-first, self-directed, experience-anchored (Knowles / andragogy).
Scenario-Based
Decisions in context with branching consequences, not content dumps.
Microlearning
Focused, spaced, ≤5-min units mapped to single objectives.
Performance Support
Job aids at the moment of need — learning that lives in the workflow.
Assessment Design
Aligned, criterion-referenced items that measure the objective.
Learning Analytics
xAPI / LMS data tied to adoption and behavior-change signals.
Accessibility
WCAG 2.2 AA, plain language, multi-device by default.
UX for Learning
Cognitive-load management, clear wayfinding, motivation by design.
Open the real thing — not slideware.
Live deployed systems
Shipped web builds — agent-supervision sim, technical tutorials, 3D training assets. Every case study opens the real system.
Source repositories
Public code on GitHub — including AeroMind FlightLab and the Gas Turbine 3D tutorial.
Interactive 3D web objects
Three.js training assets — a 360° turbine cutaway and a 42U rack — embeddable in Rise as web objects.
Human-oversight simulation
AI Workflow Quest’s approval-gate engine: deterministic branching, context risk, rewindable consequences.
Performance support at the point of work
A printable pre-use safety card with sign-off — governance thinking on one page.
Instrumentation built in
This site ships custom analytics events (project opens, outbound clicks) — measurement is part of the build, not an afterthought.
I build governed AI systems — and the human layer that makes them work.
I work in technical, high-stakes environments where trust is the product: AI workflows that keep humans over the loop, knowledge systems grounded in cited sources, and the training that turns both into adopted practice.
My edge is that I ship. I don’t stop at specs or storyboards — I build the branching engines, 3D assets, and instrumented interfaces myself, so the design intent and its guardrails survive all the way to production.
Across my work, the pattern is translation: converting complex systems into usable learning experiences for the people expected to act on them. Whether the subject is enterprise AI, jobsite safety, or decision-making under pressure, my goal is the same — help teams understand the system, trust the process, and perform with confidence.
Core Competencies
Credentials & frameworks.
Credentials & Frameworks
Let's build AI systems people can trust — and actually use.
Available for Knowledge Engineer, AI Enablement, Applied AI, and agentic-systems roles — and for learning-systems work where adoption is the hard part. Bring the governance gap or the performance gap; I'll bring the system.
Druva.akhil@gmail.com →