0
Akshat Vashishtha — Portfolio

AI · Security · Systems

Akshat Vashishtha

Builder  ·  Researcher  ·  Sophomore at Wake Tech

About

Akshat VashishthaBuilder · Raleigh, NC

While most students are still finding their footing, I'm shipping AI systems and authoring the standards that decide whether they can be trusted.

I'm a sophomore at Wake Technical Community College, studying Business Administration with an Information Technology concentration. But the credentials came before the classroom — I authored the Cloud Security Alliance's AI Controls Matrix, evaluated frontier AI models for Handshake as a Generative AI Research Fellow, and shipped Lightmus, a live deepfake-forensics tool, solo end-to-end. I build at the intersection of AI and business — with a focus on whether what gets built can actually be trusted.

I founded the AI Club — a multidisciplinary group exploring AI's impact across tech, science, the arts, and biology — serve as Secretary of the South Asian Student Association, and led a wearable-tech startup as CIO & Product Lead through TiE Young Entrepreneurs. I build because building is how I think.

0Certifications
0CSA Publications
0AI Products Shipped

Publications · CSA

Author

01 / 06
CCloud Security Alliance
AI Governance · Framework

AI Controls Matrix

Authorv1 · 2025
CCloud Security Alliance
Data Security · AI

Data Security within AI Environments

AuthorResearch
CCloud Security Alliance
Healthcare · AI

AI in Medical Research

ReviewerSep 2024
CCloud Security Alliance
Privacy · Engineering

The State of Data Privacy Engineering

ContributorResearch
CCloud Security Alliance
Offensive Security · AI

Using AI for Offensive Security

ReviewerAug 2024
CCloud Security Alliance
AI Safety · Certification

Trusted AI Safety Expert (TAISE)

ReviewerFramework
Lightmus

Featured Project · Case File — LM-0001

Founder · Shipped solo, end to end

See the evidence,
not just a verdict.

Lightmus checks images, video and writing for signs of AI generation and manipulation — then shows you exactly why, on a litmus strip you can read at a glance. No bare score to take on trust: every result comes with the proof behind it, and teaches you to spot fakes yourself.

ROLE — Founder & sole builder.
Forensics pipeline · model orchestration · product design · front-end · deploy.

000%
Strong signs
of AI generation
◂ AuthenticManipulated ▸
Analysis complete · 4 / 4 checks

Four independent checks,
one honest reading.

01 — 04 · each weighted into a single verdict

01

Where did this come from?

Provenance · EXIF / C2PA

Reads the hidden camera and app data tucked inside a file — like a receipt — plus the C2PA chain of custody when it's present.

02

Has it been edited?

Error-Level Analysis

Surfaces the compression and editing seams the eye can't see, flagging regions that were pasted, cloned, or re-rendered.

03

Does it have the tells of AI?

AI Reasoning

Catches the giveaways generative models leave behind — six-fingered hands, warped ears, impossible light and texture.

04

Have we seen this before?

Source Trace

Traces the content back toward where it first surfaced, so a familiar fake doesn't get waved through as brand new.

Four checks, not one guess
It shows its work
Honest about uncertainty
It gets sharper with use
FastAPIReactMultimodal AIImage ForensicsEXIF / C2PADocker

Web app + browser extension
Scans your feed across TikTok · Instagram · YouTube · X

Selected Work

In Progress

01 — Autonomous Agents

Vigil AI

A task-execution AI system built for true work delegation — not conversation. Five specialized agents (Orchestrator, Executor, Reasoning, Memory, and Risk) coordinate to plan, execute, and report on complex goals from a single plain-English prompt. You describe the outcome; Vigil owns every step — a background operator, not a chatbot. Targeting the $50B+ autonomous-agent market.

Node.jsMulti-AgentOrchestratorReasoning EngineMemory & RiskLocal-First

02

Oazis Armband

Co-founded a hydration-tracking wearable startup as CIO at TiE Young Entrepreneurs Regionals 2025. Sweat-sensing armband using skin-conductivity sensors + Bluetooth mobile app for real-time hydration monitoring. TAM $40–60B. Built working MVP.

Hardware · HealthTechTiE Young Entrepreneurs

03

AI Controls Matrix

Authored the Cloud Security Alliance's AI Controls Matrix — a published industry framework for governing AI systems in enterprise environments. 243 controls across 18 security domains. Recognized as a Subject Matter Expert by the CSA.

Research · Standards↗

Experience

Nov 2025 — May 2026

Generative AI Research Fellow

Handshake AI

Evaluated frontier large language models including GPT-5, Claude, and Gemini against real-world task benchmarks. Contributed to training-data curation and reward-signal design to improve model alignment and output quality. Delivered product insights that directly influenced model development priorities and deployment decisions.

Frontier LLM EvaluationTraining Data CurationReward Signal DesignProduct Insights

Aug 2024 — Jul 2025

Chief Information Officer / Product Lead

TiE Young Entrepreneurs

Led a 4-person team as CIO and Product Lead in TYE's competitive entrepreneurial program. Owned the full product roadmap for Oazis — a biosensing hydration-tracking wearable. Conducted user interviews and competitive analysis to validate the market opportunity, then pitched a polished MVP to a panel of business executives.

Product RoadmapUser InterviewsCompetitive AnalysisWearable Tech

2024 — Present

Author & Subject Matter Expert

Cloud Security Alliance

Recognized Subject Matter Expert contributing to multiple CSA working groups. Published author and reviewer across six CSA research papers covering AI governance, data security, offensive security, and medical AI.

All six publications — each with my role, proof, and the official CSA source — are showcased in the Publications wheel ↑.

AI Controls MatrixAI SafetyData SecurityPublished Author

Pitch Deck · Oazis

The deck I pitched to executives.

Oazis Armband — the hydration-tracking wearable I led as CIO & Product Lead through TiE Young Entrepreneurs. Click through the actual pitch below, or open it fullscreen.

Oazis pitch deck slide
Click slide to advance →

Credentials

Verified expertise, not just enthusiasm.

Thirteen industry certifications across AI, cybersecurity, cloud, and product — earned from CompTIA, Microsoft, IBM, ISACA, TryHackMe, and the Cloud Security Alliance. Each one earned. None inherited.

CompTIA Security+CompTIA
Microsoft AZ-900 Azure FundamentalsMicrosoft
Trusted AI Safety Expert (TAISE)CSA
Certified Information Systems Auditor (CISA)ISACA
Product Discovery / AIProduct School
Artificial Intelligence FundamentalsIBM
Cybersecurity FundamentalsIBM
SOC Level 1TryHackMe
Cloud Computing FundamentalsIBM
Cyber Threat IntelligenceIBM
Penetration Testing & Incident ResponseIBM
PreSecurityTryHackMe
Cybersecurity Roles & OS SecurityIBM

Résumé

The full picture, one page.

Every role, publication, and certification — condensed onto a single sheet. The preview is blurred; reveal it to read on-page, or open the full PDF.

↓  Download PDF
Akshat Vashishtha résumé preview (blurred)
Preview blurred

Capabilities

Engineering
  • Python
  • HTML / CSS / JS
  • Canvas API & Animation
  • Multi-Agent Architecture
  • REST APIs
  • Vercel · Railway · Render
AI / ML
  • Anthropic Claude API
  • OpenAI API
  • LangChain
  • RAG
  • Prompt Engineering
  • Model Evaluation
  • Agentic Orchestration
  • Computer Vision
  • NLP
Security
  • CompTIA Security+
  • Cloud Security (CSA)
  • HIPAA Security Rule
  • CIS Critical Controls
  • Ethical Hacking / CTF
  • SOC Operations
Product & Business
  • Product Roadmap
  • User Research
  • Customer Discovery
  • MVP Definition
  • A/B Testing
  • Wireframing (Figma)
  • GTM Strategy
  • Outbound Sales
  • Market Research
  • KPI Tracking
Tools
  • Excel / PowerPoint
  • Notion / Jira
  • Figma
  • SQL
  • Git
  • Linux
  • Azure / AWS
  • Agile / Scrum

What I'm Building

I don't wait.
I build.

Most people my age are figuring out what they want to study. I'm figuring out what I want to ship next. My focus is on agentic AI systems, security automation, and tools that do real work — not demos.

Autonomous AI security agent for real-time threat analysis
Deeper agentic orchestration with memory + reasoning
Open-source AI tooling for the security research community
Applying product thinking to AI infrastructure design

Let's connect

Build something
remarkable.

akshatvash@gmail.com