Nanotechnology Engineering
University of Waterloo · Dean’s Honours List
PetarTodorović
I build AI systems that turn complex scientific and industrial data into trusted decisions.
Palantir · Applied AI · Semiconductors · Scientific R&D
AIreasoning · learningMaterialsstructure · discoveryMarketsinformation · incentivesSystemsarchitecture · controlMove to perturb the system
01 / SYSTEMS IN PRACTICE
From data to trusted decisions.
At Palantir, I work where research, software, and consequential operations meet—turning complex data into systems that people can trust and use.
View the semiconductor whitepaper ↗Swipe to explore · 1 of 4 · Data
Data + Models selected · Secure semiconductor data ecosystem
01Semiconductor R&D · Secure data sharing
Secure semiconductor data ecosystem
Led a four-person team designing and implementing a data science module for secure data sharing, machine-learning workflows, and materials insight.
ChallengeSecure data sharing across a semiconductor ecosystem.OwnershipDesign and technical implementation of the data science module.SystemMachine-learning and statistics-based workflows.4-person teamRead whitepaper ↗ 03Renewable technology · Solar cellsManufacturing root-cause intelligence
Applied LLM-driven reasoning to diagnose production failures and surface actionable insight across complex, multi-stage solar-cell fabrication workflows.
ChallengeProduction failures across multi-stage fabrication workflows.MethodAdvanced agentic reasoning and LLM-driven root-cause analysis.Public proofRenewable-technology DevCon talk.LLM reasoningWatch DevCon ↗ 04Consumer packaged goods · Supply chain managementEnterprise supply-chain transformation
Owned end-to-end development of five supply-chain workflows, partnering with strategic leadership across multiple high-impact initiatives and carrying delivery from a bridge-financed pilot through enterprise adoption.
ConversionSolely drove the technical and implementation work that converted the pilot into a two-year, enterprise-wide agreement valued at $X million.ExecutionDesigned scalable workflows and led execution across complex IT constraints, external clients, engineers, and stakeholders.ExpansionExtended the account through multiple build and agent camps while managing and mentoring three junior engineers.5 workflows3 engineers mentoredWatch AIP Con 5 ↗ 05Product strategy · Modeling infrastructureModel Studio: from customer need to platform capability
Served as the primary link between product and business development, owning the product vision end to end and translating business requirements into technical deliverables.
Product ownershipConverted business deliverables into engineering requirements and carried the product vision across a cross-functional initiative.Reusable designDrove stakeholder engagement around real customer needs, with generalizability as a core principle to maximize reuse across use cases.Production outcomeDelivered the product to production; it is now adopted as an official microservice in Palantir’s modeling arsenal.Product → engineeringProduction microserviceExplore Model Studio ↗02 / RESEARCH
Research with a path to practice.
Published research, industry work, and open tools across interpretable machine learning, semiconductors, and optical materials.
Published research · 2023Industry whitepaperOpen-source tool
DARWIN
Interpretable discovery of semiconductors with machine learning · npj Computational Materials
Accelerating Semiconductor R&D
Connected data and workflows for semiconductor research and development · Palantir
Luminescence Profile Generator
Ideal luminescence curve generation for optical materials research · GitHub
03 / LEADERSHIP
Technical depth, made operational.
I lead at the intersection of scientific rigor, executive priorities, and the day-to-day work required to move systems into practice.
- 04Lead technical delivery
Leading a four-member team designing and implementing a semiconductor data science module.
- 30Scale systems into use
Agentic R&D infrastructure active across 30 deployments and serving 300+ users.
- ↗Connect strategy to implementation
Scope business use cases with executive leadership across structured and unstructured datasets.
- R&DTranslate technical work
Write industry whitepapers and technical materials for semiconductor participants and external teams.
04 / APPROACH
From first principles to real-world systems.
My work moves between scientific research and consequential deployment: reducing complex problems to their fundamentals, building rigorous models, then engineering systems that create durable value.
?Problem
△Model
⌘Build
↻Learn
05 / FOUNDATION
Scientific foundations, applied.
Electrical Engineering
University of Toronto · Perovskite nanomaterials
Deployment Strategist / Senior Data Scientist
Palantir Technologies · Applied AI in practice
The most interesting problems are still unsolved.
Let’s build what’s next.
For conversations about applied AI, scientific R&D, and complex decision systems, LinkedIn is the direct path.
Connect on LinkedIn