Manifesto
For years, I've been drawn to the intersection of technology and human potential, from enterprise AI systems to grassroots empowerment. I believe in building with purpose — not just what works, but what matters. In asking better questions, before designing better answers. In seeing systems where others see screens, and stories where others see specs.
Empathy is my method. Curiosity is my fuel. Systems thinking is my lens. And clarity is my craft.
This portfolio is not just what I've done —
It's how I think.
It's why I build.
It's my climb of curiosity.
Harsh Kataria
Climb of Curiosity
"Ask me how I scaled an AI SaaS product 5x across 17,000+ devices."
Led end-to-end development of Cisco's AI SaaS platform enabling predictive insights for service providers. Drove product lifecycle, defined requirements through 30+ stakeholder interviews, and built analytics dashboards tracking 15+ KPIs to improve engagement and satisfaction.
Tools/Skills: Python, REST APIs, Data Visualization, Product Analytics, Agile
Outcome: 5x adoption growth, 280+ hours saved per deliverable, 25% reduction in support queries
"Wanna know how I aligned product, pricing, and GTM?"
Created data-driven roadmaps for B2B and B2C product lines, synthesizing insights from competitive research and customer analysis. Designed pricing and marketing frameworks informing business strategy for a $2M+ property management system.
Tools/Skills: Market Research, Product Roadmapping, Pricing Strategy, Competitive Analysis
Outcome: Projected 15% revenue improvement across 3 market segments
"Ask me how I tested LLMs and Bayesian models to optimize restaurant inventories."
Built early-stage AI prototypes integrating multimodal reasoning to improve SMB inventory decisions. Conducted GTM validation, P&L modeling, and data storytelling dashboards for strategic pivots, leading to measurable engagement and positioning insights.
Tools/Skills: Python, Bayesian Analysis, LLMs, Meta Ads, A/B Testing, Data Storytelling
Outcome: 15% higher engagement from optimized GTM strategy
"Ask me how I turned ML research into a $200K+ business case for predictive maintenance."
Drove AI productization of predictive maintenance for Honda's R&D wind tunnel by translating research models into business value. Bridged technical and operational teams to validate data pipelines and align ROI-driven metrics.
Tools/Skills: Applied ML, Business Modelling, Stakeholder Management, Strategy Design
Outcome: $200K+ annual savings projected for $124M test facility
"Ask me how I led a 5-member team to build an enterprise-grade code reasoning AI-as-a-service."
Managed cross-functional delivery of an AI-as-a-service product for enterprise code reasoning. Owned roadmap, backlog prioritization, and stakeholder syncs aligning scientists and engineers toward a 15-week release cycle.
Tools/Skills: Agile, LLM Evaluation, TypeScript, Sprint Planning, Product Management
Outcome: Projected 30% improvement in developer productivity
Graduate program blending product strategy, metrics, futures thinking, and financial modeling.
Conducted deep learning and NLP research, published papers with ACL, and filed a patent for thought-to-text AI.
Awarded for driving 30% reduction in customer support tickets and improving cross-team collaboration.
Recognized for innovation excellence, customer satisfaction (CSAT 4.8/5), and community leadership through 'Delhi Talks Tech'.
Among 200+ teams tackling UN Sustainable Development Goals with social impact solutions.
Secured patent for AI-powered assistive mobility interface using EEG signals.
Patented EEG-based system translating neural signals into text for assistive mobility applications.
View Patent →Implemented hierarchical ML models achieving 6% F1 improvement in legal document processing.
View Paper →Developed hybrid stacked-embedding model for toxic span detection, improving NLP classification accuracy.
View Paper →