Skip to main content
ABOUT SHIVA SHANKAR S

I turn AI concepts into systems that can operate in the real world.

I am an AI Engineer focused on building systems that go beyond conversational demos. My work spans retrieval-augmented generation, tool-calling agents, intelligent automation, voice interfaces, AI evaluation, safety guardrails, and production-grade full-stack development.

I am especially interested in the boundary between probabilistic models and deterministic software. When accuracy matters, I prefer explicit tools, validated calculations, structured outputs, measurable evaluation, and human escalation rather than relying on a model to improvise.

I currently build from Bengaluru, India, and I am working toward contributing to ambitious AI teams in Spain and the broader European technology ecosystem.

“Good AI engineering is not about making a model sound intelligent. It is about designing a system that behaves reliably.”

12.9716° N, 77.5946° ESYS // 2026
Shiva Shankar S — AI Engineer
SHIVA SHANKAR S
Career Direction: Bengaluru, India to SpainAbstract geographic coordinate route illustrating Shiva's ongoing work in Bengaluru and clear long-term direction toward Spain and the European AI ecosystem.Bengaluru, IndiaBUILDING NOWINTERNATIONAL ↗Spain / EuropeLONG-TERM DIRECTION
RESEARCH

Exploring targeted machine unlearning for responsible LLM systems.

IEEE IES GENERATIVE AI CHALLENGE — FINALIST PAPER

Surgical Amnesia: A Framework for Targeted Machine Unlearning and Copyright Compliance in Large Language Models

Co-authored a parameter-efficient machine unlearning framework using Llama 3.2 1B and LoRA with a negated cross-entropy objective. The work explored targeted data removal without full model retraining and considered obligations connected to data erasure, copyright compliance, GDPR Article 17, and the EU AI Act.

IEEE Challenge Outcome: The project reached the final round of the IEEE Industrial Electronics Society Generative AI Challenge. Revised methodology per reviewer feedback and submitted in camera-ready IEEE conference format.

Research Focus Vectors

  • Parameter-Efficient LoRA Unlearning
  • Negated Cross-Entropy Objective
  • GDPR “Right to be Forgotten” Compliance
  • EU AI Act Data Erasure Alignment
  • Model Utility vs Forget Efficacy Trade-off
RESEARCH PIPELINE // IEEE GENAI CHALLENGE FINALIST

Surgical Amnesia Machine Unlearning Architecture

IEEE IES Camera-Ready Format
1. BASE MODEL
Llama 3.2 1B
Pre-trained Base LLM
2. TARGETED FORGET SET
Data Erasure Target
Copyright / GDPR Set
3. LoRA UNLEARNING
LoRA Adapter Layer
Negated Cross-Entropy
4. UPDATED MODEL
Unlearned Model
Retained General Reasoning
COMPLIANCE & UTILITY EVALUATION HARNESS
Forget Efficacy Score
Verifies erased information cannot be reconstructed or extracted via prompt probing.
Model Utility Retention Score
Ensures general knowledge, language coherence, and task performance remain intact.
JOURNEY

Building toward a career in applied AI engineering.

2022 – 2025Kristu Jayanti (Deemed to be University), Bengaluru

Bachelor of Computer Applications (BCA)

Completed undergraduate degree in computer applications. Built practical experience in software development, web engineering, databases, and big data analytics coursework (Hadoop, MapReduce, NoSQL, Hive, Pig).

2025 – 2027St Joseph's University, Bengaluru

Master of Computer Applications (MCA)

Currently pursuing Master's degree with dedicated focus on AI/ML engineering, RAG systems, LangGraph agentic architectures, tool-calling systems, and LLM evaluation harnesses.

2025 – 2026

Transitioned from full-stack development into AI systems engineering

Shifted focus from conventional web backends to architecting production-grade AI systems combining probabilistic language models, deterministic software tools, vector databases, and safety guardrails.

2026

Built AthleteIQ, AI Receptionist, Alpha Lead Engine, Reframe & Smooth Operator

Designed and shipped 5 complete production AI systems featuring LangGraph ReAct agents, real-time voice automation, 6-module sales pipelines, client-side encryption, and dating RAG coaching.

2026IEEE Industrial Electronics Society

Reached IEEE Generative AI Challenge Final Round

Co-authored 'Surgical Amnesia: A Framework for Targeted Machine Unlearning and Copyright Compliance in Large Language Models' using Llama 3.2 1B and LoRA with negated cross-entropy.

Current DirectionBengaluru → Spain / European Tech Ecosystem

Building from India toward AI engineering opportunities in Spain

Actively positioning for AI Engineer roles in Spain and Europe, combining production-focused system reliability, applied AI research, and European technology ecosystem alignment.

Target Market Focus: Spain & Wider European AI Ecosystem
BEYOND ENGINEERING

Leadership, competition, and building beyond the classroom.

While AI engineering is my primary focus, many of the skills I rely on today were developed long before I began building intelligent systems. Sport taught me discipline, teamwork, and resilience. Student leadership taught me responsibility and communication. Technical competitions taught me to solve unfamiliar problems under pressure. Independent client work taught me how to turn real business challenges into practical solutions.

LEADERSHIP

Sports Leadership

Leadership through sport

I currently serve as the Sports Representative for the MCA Department at St. Joseph's University. During undergraduate studies at Kristu Jayanti University, I was elected Sports Secretary, representing students and organizing major events. Earlier, I served as my School Sports Captain and captained my college football team.

KEY ROLES

  • Sports Representative, MCA – St. Joseph's University
  • Former Undergraduate Sports Secretary
  • Captain, Undergraduate College Football Team
  • Former School Sports Captain
COMPETITION

Competitive Football

Competing as a team

Football has been one of the biggest influences on how I approach leadership and teamwork. I represented Bangalore Yellows in the Bangalore football league and was part of the team's promotion from the C Division to the B Division. Playing in a competitive environment reinforced consistency, preparation, adaptability, and trust.

Bangalore Football LeagueManchester United Fan

IEEE Research Milestone

IEEE Industrial Electronics Society Generative AI Challenge

TOP 25 WORLDWIDE

Our research paper, “Surgical Amnesia: A Framework for Targeted Machine Unlearning and Copyright Compliance in Large Language Models,” explored parameter-efficient machine unlearning using Llama 3.2 1B and LoRA, focusing on responsible AI, targeted knowledge removal, GDPR compliance, and the European AI Act.

Global Finalist — Australia Conference Finals

Selected among the Top 25 teams globally and invited to present at the conference finals in Australia. Reaching the final stage of an international IEEE challenge reinforced my commitment to trustworthy AI and machine learning reliability.

HACKATHONS

Hackathons & Technical Competitions

Building under pressure

I actively participate in hackathons and inter-collegiate technology competitions. Competing across institutions including Christ University and Kristu Jayanti University has sharpened my ability to solve practical challenges, collaborate with diverse teams, and execute under tight deadlines.

ORGANIZATION

Campus Involvement

Beyond academics

Throughout college, I have contributed to organizing technical festivals, sports events, and inter-collegiate activities. Event coordination, planning, and student engagement taught me that successful outcomes depend as much on organization and communication as on technical ability.

Independent Client Work

Building for real clients

REAL BUSINESS IMPACT

Alongside academics, I work independently building websites, AI automation systems, and digital products for clients. Working directly with businesses taught me to understand requirements, iterate rapidly based on feedback, and deliver practical solutions rather than academic exercises.

AI-powered business automation
Website design & development
Workflow automation
AI agents & intelligent systems
Full-stack product development
Technical consulting
CORE PHILOSOPHY

“Understand the objective. Take ownership. Work well with others. Deliver meaningful results.”

Whether I am leading a football team, representing students, contributing to research, competing in a hackathon, organizing an event, or building software for a client, I approach every challenge with the same mindset. Those experiences continue to shape how I build AI systems—reliable, practical, and designed for real-world impact.

NOW

What I am building and learning now.

Building

  • Modular agent workflows in LangGraph
  • Production AI evaluation & assertion harnesses
  • AI-powered outbound sales automation (Alpha Lead)
  • Reliable tool-calling applications with safety bypass

Learning

  • Multi-agent LangGraph patterns & supervisor state
  • Langfuse tracing, latency, & observability
  • Docker & containerized production deployment
  • Google Cloud Run serverless microservices
  • MLOps foundations & production model drift

Direction

  • AI Engineering & Applied ML roles
  • Applied AI, LLM systems, & RAG architecture
  • Remote opportunities & relocation to Spain
  • Building systems with measurable real-world value
CONTACT

Let’s build AI systems that create real value.

I am open to conversations about AI Engineering, applied AI products, LLM applications, intelligent automation, remote collaboration, and opportunities in Spain.

Bengaluru, India · Open to remote work and relocation to Spain
Bengaluru, IndiaBUILDINGSpain / EuropeRELOCATING LONG TERMCOLLABORATE ↗