profile-pic

Somil Singh

I’m a Software Engineer at Oracle by day, building scalable enterprise systems where precision is everything. But my mind is usually orbiting a bit higher.

My ultimate goal? To engineer the AI that will guide the next generation of space exploration. I’m obsessed with creating systems that can think for themselves when they are millions of miles away from home, namely Autonomous Systems for space exploration.

My Technical Playground:

• The Engineer: ~2 years of shipping production-grade code and solving complex problems at scale.

• Build Fast: Learnt React.js in 3 days to deliver a critical Full-Stack AI feature in under a week.

• The Researcher: Published author on object tracking systems and former researcher at IISc Bangalore, constantly looking for ways to make algorithms "learn" like we do.

If you’re a recruiter looking for an engineer obsessed with what’s next or just someone who loves Interstellar as much as I do, send me a connection request. Let's chat. 🌠

  • Role

    Full-Stack AI Engineer

  • Years of Experience

    4.3 years

  • Professional Portfolio

    View here

Skillsets

  • CNNs
  • World Models
  • Vlm
  • Vla
  • Vision Transformers
  • Vector DB
  • Rest APIs
  • model quantization
  • Model evaluation
  • LLMs
  • JEPA
  • fine-tuning
  • Diffusion models
  • Java
  • Transformers
  • MATLAB
  • HuggingFace
  • CUDA
  • C
  • TensorFlow
  • rag
  • PyTorch
  • Python
  • LangChain
  • JavaScript

Professional Summary

4.3Years
  • Oct, 2026 - Present 1 month

    Founder & Creator

    Inside AI
  • Jan, 2024 - Present2 yr 10 months

    Full-Stack AI Engineer

    Oracle
  • Jul, 2026 - Sep, 2026 2 months

    Founder

    Stealth AI Startup
  • May, 2023 - Oct, 2023 5 months

    Computer Vision Research Intern

    Samsung R&D Institute India
  • Jul, 2023 - Dec, 2023 5 months

    Deep Learning Research Assistant

    Iisc Bangalore
  • Jun, 2026 - Jun, 2026

    Founder

    BehaviorAI (v1)
  • Nov, 2022 - Mar, 2023 4 months

    Deep Learning Research Assistant

    Wipro Pari

Work History

4.3Years

Founder & Creator

Inside AI
Oct, 2026 - Present 1 month

Full-Stack AI Engineer

Oracle
Jan, 2024 - Present2 yr 10 months
    Engineered a multi-agent LLM ecosystem (MCP, PL/SQL DB, Jira, Codebase) that autonomously triages 150-200 bugs per week; cut resolution time from 45 min to under 2 min and saved 120+ engineering hours monthly. Developed RAG pipelines processing 500+ RFP documents daily across a 100k+ vector knowledge base using LangChain and Vector 23ai Oracle DB, achieving sub-1.5s retrieval latency and a 35% throughput increase. Spearheaded org-wide adoption of Agentic AI tools (Cline, Kilo Code, Codex) to modernize enterprise SDLC into platform-agnostic AI infrastructure, generating $1.1M in projected savings. Owned 19 full-stack features across React and Java-Spring in OPC, improving responsiveness by 28%; resolved 40+ high-priority bugs with 0-defect post-fix record.

Founder

Stealth AI Startup
Jul, 2026 - Sep, 2026 2 months

Founder

BehaviorAI (v1)
Jun, 2026 - Jun, 2026

Deep Learning Research Assistant

Iisc Bangalore
Jul, 2023 - Dec, 2023 5 months
    Researched Neural Radiance Fields (NeRF) using Plenoptic Functions to model 7D scene representations for human-pose estimation, boosting reconstruction fidelity by 25% on benchmarks; validated on Blender, LLFF, DTU. Integrated SfM/SLAM camera calibration for 3D scene reconstruction with differentiable ray tracing, maintaining real-time inference at 12 fps; designed end-to-end rendering and spatial data pipelines in PyTorch on Ubuntu. Mitigated visual aliasing using Gaussian splatting and supersampling via custom acceleration structures (BVH), reducing mean joint-position error by 18% on out-of-distribution (OOD) datasets.

Computer Vision Research Intern

Samsung R&D Institute India
May, 2023 - Oct, 2023 5 months

Deep Learning Research Assistant

Wipro Pari
Nov, 2022 - Mar, 2023 4 months
    Trained a customized Single Shot Detector (SSD) with Feature Pyramid Networks (FPN) for multi-scale perception across dense autonomous driving scenarios, achieving 55-75% mAP on real-world traffic datasets (WIRIN).

Major Projects

1Projects

Object Detection, Classification & Tracking of Everyday Common Objects

    Built GPU-accelerated object segmentation and tracking pipeline (YOLOv4, TensorFlow, OpenCV) with sub-30ms latency; combined detection with dynamic cropping for 75-95% accuracy, handles edge-frame targets and occlusion.

Education

  • B.E in CSE

    Rashtreeya Vidyalaya College of Engineering (2024)
  • High School Diploma

    National Public School, Indiranagar (2020)