Dheeraj Santosh Kollipara — Lzy Lad
Local AI Setup & On-Device AI Consultant
nagadheeraj.kollipara@gmail.com
+91 89856 21250
linkedin.com/in/lzylad
behance.net/lzy_lad
Guntur, Andhra Pradesh, India
Profile
I run my entire AI stack on local hardware — RTX 4050, 6GB VRAM — with ComfyUI, Ollama and local transcription pipelines that don't depend on external API availability. I set that up for other people: models, workflows, agents and cost-tuning on their own machines. I understand inference economics from the inside, which means my advice carries no vendor incentives.
Proof of Work
- ~40% internal process automation, Jagriti Yatra (700 riders, core team of 7)
- 90+ AR/VR/Web3/UX/AI workshops + XR Summit IIT Madras with Steve LaValle
Experience
2024
Generative AI Trainer / AI Consultant
Integral World, Auroville + institutional FDP engagements
- Applied AI workflows, automation and digital tooling for real operational use, plus role-specific usage guidance for non-technical staff.
2012 – present
Founder & Creative Technologist (Freelance)
Lzy Lad / Lzy Labs
- 14+ years across design, graphics, media, art, technology, science and AI, delivering creative solutions and strategies.
- Brand identity, creative campaigns (static, motion, 360°), generative/psychedelic art, VJ sets, projection mapping, live AR painting for concerts.
2022
Core Team — Design, Communications & Programming
Jagriti Yatra (700-participant national train journey)
- Automated roughly 40% of internal operational process across a distributed national team.
2019 – present
Workshop Facilitator — AR/VR, Web3, UX & AI
Pan-India (college faculty, students, creative professionals, working teams)
- 90+ workshops and sessions covering AI workflows, AR/VR, blockchain fundamentals and UX design.
- Formats from half-day to three-day intensive laboratories; live demonstration and participant-built artifacts throughout.
Skills & Tools
Local Multimodal Pipelines
- ComfyUI, Stable Diffusion, diffusion video
- Local open-source workflows for text, image, video, audio and 3D
- Text-to-speech and speech-to-text pipelines
- Vector graphics, web design, audio, 3D and Web Experience Design
- Speech-to-text (faster-whisper, GPU), audio spectral analysis
- 3D Gaussian Splatting, NeRF (training, editing)
- TouchDesigner, Meta Quest 3, WebXR
Generative AI & Local Models
- ChatGPT, Claude, Gemini, Perplexity, DeepSeek (daily working instruments)
- Advanced prompt engineering, context engineering, CoT, meta-prompting
- Output evaluation, fact verification, hallucination detection
- Ollama (local models, PC and mobile, text generation)
Engineering
- Claude Code, Cursor, GitHub Copilot, Hermes (AI-assisted product building)
- Vibe coding: Cursor, Bolt, Lovable, v0, Replit
- AI-assisted web design, AI-assisted graphic design
- AI-assisted automation
- MCP tool servers, loop engineering, graph engineering
Automation & Orchestration
- AI-assisted data cleaning & interpretation, data visualisation/dashboards
- Multi-agent research systems, multi-agent orchestration
- Agentic orchestration architecture & handover (consulting, not running)
- Browser automation via CDP, MCP tool servers
Audio & Voice ML
- faster-whisper (local GPU transcription), voice agents
- Audio spectral analysis, audio-reactive visuals
- Sound design, soundscaping
Education & Logistics
B.E. Civil & Structural Engineering — Sri Chandrasekharendra Saraswathi Viswa Mahavidyalaya (SCSVMV), Kanchipuram, Tamil Nadu. 2012–2016, 8.3/10 aggregate.
English, Telugu, Tamil (functional).
Remote worldwide, or in-person anywhere in India. Flexible hours.
Lab: RTX 4050 (6GB VRAM) workstation with local model execution — labs and pipelines do not depend on external API availability.