pranav

~ ❯ whoami

pranav jha

applied ai · full-stack · game dev

building

▸ Mebius — an AI that points you through any screen

notable

learning

agent memory, evals and computer-use agents

reading

ai engineering · chip huyen

today
0
this week
137
streak
5d

contributions to public repos

languages across repos

  • TypeScript57%
  • CSS14%
  • C#7%
human
100 commits
ai-assisted
214 commits

↻ GitHub activity, refreshed hourly. ai-assisted = commits over the last 30 days co-authored by an AI tool.

~/pranav ❯ whoami

pranav jha

I'm an Applied AI Engineer who is passionate about building AI products that solve real problems and make people's work simpler.

▸ building Mebius — an AI that points you through any screen

  • python
  • typescript
  • next.js
  • nestjs
  • postgresql
  • gemini
  • langchain
  • rag
  • gcp
  • docker
  • unreal engine 5
  • c++
  • unity
contributions
800
active days
68
longest streak
8d
best day
57

hover a day

Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
@PranavJha-3000 ↗lessmore

❯ experience

  1. founder — upstagexmar 2026 – sep 2026

    Built creator discovery for brands: a 10 lakh+ vetted creator dataset, fit-scoring and pricing workflows, and a discovery → AI brief → contract → escrow → payout flow on Next.js, NestJS, PostgreSQL, GCP and Gemini. Reached 15 brands, 30 creators and INR 5L+ processed.

  2. unreal developer, freelance — app mechanicaug 2025 – sep 2025

    Turned a short-drama concept into Microdose, a playable game-to-episode loop, and carried it alone from concept to Android build.

  3. software developer intern — idz digitalmar 2025 – jun 2025

    Scoped and shipped UE5 features: gameplay, procedural generation and object pooling. Profiling-led fixes brought 30% faster loads and 20% better frame stability.

❯ education

  1. b.tech, electronics & communication — iiit ranchi2023 – 2027 (expected)

    CGPA 8.00. Coordinated Arcanum, the game dev & design club: grew it to 100 active members, ran workshops for 500+ students, and hosted Dungeon Dev, a 72-hour national game jam for 500 participants.

  2. aissce (pcm) — cambridge school, indirapuram2022

    Class XII with physics, chemistry and maths. Scored 90%.

❯ about

I like taking a vague problem and walking it all the way to working software: talk to the people who have it, scope it, build it, measure it, fix it.

That has meant AI and data workflows, full-stack products and real-time systems, across creator tech, desktop AI, Android and Unreal Engine.

I'm studying Electronics & Communication at IIIT Ranchi (graduating 2027), and I'm at my best in the zero-to-one stretch.

❯ contact