I compose, perform, model in Blender, direct short films, and ship production AI software. Most candidates have one of these. The work below is the proof.
Open jashtechno.com/coffee-shop
I composed and produced the music. I performed in it. Every object — chairs, lamps, cups, the whole scene — modeled and animated from scratch in Blender. Then I integrated the assets onto a Three.js / WebGL web app and made the whole experience playable. Users interact with the music and the scene together.
One person. One pipeline. Real shipped work.
Music direction for an Indian web series — composition, song direction, and audio production decisions across episodes.
Independent composition and production. Mixed and mastered solo.
Independent composition and production. Mixed and mastered solo.
I'm a music composer, lyricist, and audio engineer — I mix and master my own productions. The reason this matters in software-and-AI contexts: I know what audio post sounds like when it's right, what a director cares about in a recording session, and how scoring decisions interact with picture edit.
Production AI platform deployed to a paying customer running a 300+ property multi-entity portfolio. 73 versions of customer-driven iteration over the past year. Multi-API fallback chains, structural validators, audit logging with replay.
Autonomous lead-to-call workflow: Gmail → OpenAI → web scraping → OpenAI → Google Calendar → Vapi → Twilio voice agent. Two sequential LLM calls with structured JSON parsing, fuzzy address normalization, end-to-end orchestration. Same shape as multi-node generative video pipelines.
Real-time analytics dashboard with React, WebSockets, complex chart UIs, full WCAG-compliant component library, end-to-end CI/CD with Jest and Cypress.
I direct, shoot, edit, color grade in DaVinci Resolve, and finish my own short films — from script and storyboard through final render.
NLE editing, sound design, color, final delivery. Lived experience of what artists and directors actually care about during production — where AI tools meaningfully help vs. where they get in the way.
Which means when I look at a generative video model dropping into a partner's existing After Effects / Premiere / Houdini / Nuke pipeline, I'm not theorizing about where the friction lives. I've been the person staring at a render queue waiting for a take that didn't quite work.
I'm Palleti Jashwanth — Jas. I run JashTechno, my independent software studio, alongside a senior frontend contract at Verizon (TypeScript, React, Next.js — led an internal portal rebuild that dropped page loads ~30%).
Before Verizon, I spent 3+ years at State Street Corporation building real-time, data-heavy financial web applications. M.S. Computer Science from Rowan University (May 2025). 5+ years total in production software engineering, plus the creative work above.
The combination of creative production craft (music, Blender, film) and production-grade AI engineering (LLM systems with paying customers, multi-step agentic pipelines, AWS) is what I'm building toward — and the kind of role I'm looking for.