/resume Curriculum Vitae · 2026

Jitesh Gosar.

GenAI and ML engineer with eight years across Japan and India. Agentic LLM pipelines, RAG, fine-tuning, and prompt engineering, on a deep computer-vision and synthetic-data foundation. IIT Bombay, JLPT N2, currently a Technical Specialist at Sony India Software Center.

Download PDF ↓ Get in touch ↗
01 Experience

Eight years, four teams, two countries.

2023 – Present
Technical Specialist
Sony India Software Center
Bangalore, India
  • Architecting agentic LLM pipelines for internal AI systems: speech-to-text, RAG over domain knowledge, and multi-agent reasoning orchestrated with LangChain.
  • Working on computer vision and ML systems with engineering teams across Japan and India, including synthetic data pipelines and ML evaluation workflows for camera and imaging systems.
  • Cross-cultural technical coordination as a JLPT N2 speaker, translating requirements between Japanese stakeholders and India-based engineering teams.
Agentic LLMs · RAG LangChain Computer Vision Synthetic Data JLPT N2
2022 – 2023
AI Technical Lead
Honda Research Institute · via L&T Technology Services
Tokyo, Japan
  • Led development of three applications on the Haru social robot: an emotionally-intelligent Telegram remote built with LoRA fine-tuning on a Hugging Face model, an international classroom mediator (Japan ↔ Australia), and an NLP collaborative storyteller. Deployed at 10+ research institutions globally.
  • Built the virtual simulation world for Haru from scratch in Unity, enabling safe scenario testing before real-world deployment.
  • Coordinated a 5–20 person distributed team across Japan, Spain, and Australia, aligning requirements, delivery cadence, and technical integration across time zones.
  • Educational mediator demo led to published academic research papers on human-robot interaction.
LoRA · Hugging Face Social Robotics NLP · Dialogflow Unity · React
2020 – 2021
AI Engineer / Project Lead
Pasona Tech, Inc.
Osaka, Japan
  • Led a 3-person team to build a CAD symbol and text detection system for automated construction cost estimation. Achieved 85% accuracy across 30+ classes using Mask R-CNN and keypoint models.
  • Independently created synthetic training datasets using Blender, Unreal Engine, and OpenCV, eliminating manual annotation cost.
  • Built a reinforcement learning simulation using Unity ML Agents in a 3D environment for autonomous robot behavior testing.
  • Visited customer locations to instruct on synthetic data creation methods.
Computer Vision PyTorch Synthetic Data Reinforcement Learning
2017 – 2020
Research Consultant · Autonomous R&D
Hino Motors · via CBS Techno Co. Ltd.
Tokyo, Japan
  • Contributed to the Level-4 autonomous truck program as part of a 17-person R&D team. The truck eventually drove driverlessly on Hino's real test course.
  • Led a 4-month autonomous driving demo project on a ZMP Robocar platform, mentoring new engineers and coordinating parallel development of A* path planning, YOLO object detection, PilotNet lane control, and obstacle avoidance.
  • Served as bilingual technical bridge between Japanese engineers and English-speaking university partners.
  • Established a documentation methodology for managing source code from multiple service providers.
Autonomous Vehicles Sensor Fusion A* · YOLO · PilotNet ROS
Summer 2016
Electronics Engineer · Internship
Tech Mahindra Ltd.
Pune, India
  • Led development of a Li-Fi prototype, transforming LED lamps into data transmitters using BeagleBone Black and Arduino.
02 Stack

The tools I reach for, by frequency.

GenAI & LLMs
  • OpenAI · Claude · Gemini
  • LangChain · multi-agent systems
  • RAG · vector databases
  • Prompt engineering · fine-tuning (LoRA)
ML & Computer Vision
  • Python · PyTorch · TensorFlow / Keras
  • NLP · Hugging Face · Whisper
  • OpenCV · Mask R-CNN · YOLO
  • Detection, segmentation, synthetic data
Simulation & Robotics
  • ROS · Gazebo
  • Unity (C#) · Unity ML Agents
  • Unreal Engine · Blender
  • Sensor fusion (LiDAR + camera)
Build & Ship
  • FastAPI · Flask · Django
  • React · SQL / NoSQL
  • AWS · Git · CI/CD · Docker
  • Cross-cultural delivery
03 Education & Certifications
2013 – 2017

Indian Institute of Technology, Bombay

B.Tech, Electrical & Electronics Engineering · Mumbai, India

Built and published Android apps on Google Play. Led a hackathon team building an AR app with Unity3D and Vuforia. Marketing Coordinator at Abhyuday (social festival), Organizer at Mood Indigo (cultural festival).

2023IBM DevOps and Software Engineering Specialization
2023Google Project Management Certificate (in progress)
2021JLPT N2 · 日本語能力試験
2021AWS Certified Cloud Practitioner
2019Deep Learning Specialization · DeepLearning.AI
04 Languages
English
Fluent · TOEIC 905
日本語
JLPT N2 · Full Professional
हिन्दी
Native