Skip to content

Jyoti Pokhrel / AI Engineer, Researcher

Applied AI researcher who turns open questions into working systems: agentic AI, RAG, LLM evals and multimodal models.

About

Full-stack AI engineer working across retrieval, computer vision and workflow automation.

The thread is practical research: take a paper-shaped idea, test the system boundary, then ship the smallest reliable product surface.

Education

B.Sc. Computer Science and Information Technology

Soch College of IT, Pokhara

7th Semester, GPA 3.4/4.0, Batch Topper (2nd and 3rd Sem)

Experience

Applied AI Research Intern

Ninebar

Singapore, Remote

  • 5 proofs of concept in 6 weeks, each taken from an open research question to a working demo.
  • Benchmarked 4 graph and semantic store setups on identical, checksum-frozen inputs, with a lossless round-trip test as the pass gate.
  • Natural-language to Cypher agent that checks every generated query against the live schema before running it read-only. Judged on 75 expert-written questions.
  • Vision-LLM chart extraction that keeps page text on-device, sends only chart crops to the model, and checks each series against its printed average within 2%.

AI Engineer Trainee

SkinPal AI

Nepal, Remote

  • Acne detection CV model with real-time inference on user photos. OpenCV.
  • TikTok content pipeline: trending video discovery, script generation, auto-captioning.
  • Reddit AI agent for zero-touch community outreach. Identifies threads, drafts contextual responses.

AI Fellow

Fusemachines

Nepal, Remote

  • 10+ end-to-end ML projects across computer vision, generative modeling and signal processing.
  • MLflow on every project, from tracked experiment runs to deployment-ready models.
Certificate ↗

Selected Work

Each one can be inspected: a paper, a repository, or a measured result.

Research & Writing

Skills

01

Core

  • Python
  • SQL
  • Git

02

AI / LLM

  • LangChain
  • LlamaIndex
  • RAG
  • pgvector
  • PyTorch
  • scikit-learn

03

Backend & Data

  • FastAPI
  • PostgreSQL
  • MongoDB
  • Docker
  • MLflow

04

Infra

  • Linux
  • GitHub Actions
  • AWS/GCP

Recognition

Contact

Open to AI engineering and research roles.

Email is the best place to start.

[email protected]