Thabo profile photo thabo.mphuthi

Web Resume

Thabo Mphuthi

Applied AI Engineer & Technical Leader building enterprise AI systems. Over 13 years shipping production AI-enabled software across 30+ organisations, turning advances in AI platforms into production AI products through strong engineering, rapid experimentation, and scalable platform design. Led multidisciplinary teams delivering systems at enterprise and consumer scale. Experienced in coaching engineers, building capability from the ground up, and turning field learnings into reusable platform and product improvements.

Career Snapshot

Experience

13+ years

Engineering leadership and delivery

Organisations

30+ served

Across enterprise and venture contexts

Global delivery

18 countries

Multi-market product and engineering execution

Scale

30K+ B2B 1M+ B2C

Users of AI systems and platforms built and/or led by me

Core focus

Enterprise AI delivery Applied AI/ML Platform and product strategy Engineering team leadership

Work Experience

Senior Principal AI Engineering Lead | Associate Partner

QuantumBlack, AI by McKinsey

2026 - Present

  • Lead engineering teams across multiple customer programmes, delivering production agentic AI systems from problem framing and technical discovery through implementation, deployment, and operational support.
  • Design scalable multi-agent architectures, while coaching engineers, reviewing technical decisions, and helping resolve complex production issues.
  • Led the end-to-end delivery of AI-powered workflow automation for a global media organisation,reducing procure-to-pay cycle times by more than 50%.

Principal Applied AI Engineer

QuantumBlack, AI by McKinsey

2023 - 2026 · London, UK

  • Led engineering for McKinsey's enterprise generative AI platform (Lilli) serving more than 20,000 consultants, reducing research workflows from hours to minutes through production AI capabilities.
  • Built and shipped retrieval-augmented AI capabilities integrating enterprise knowledge with frontier language models to deliver grounded, citation-backed answers under enterprise reliability and security constraints.
  • Led forward-deployed engineering teams shipping QuantumBlack's agentic orchestration platform, taking capabilities from technical discovery and architecture through client deployment, while translating field learnings and customer feedback into reusable platform capabilities and product improvements.

Principal AI & Software Engineer

McKinsey & Company, Digital Labs

2020 - 2022 · London, UK

  • Led engineering for multiple AI-enabled digital ventures across industries, partnering with executive stakeholders to shape opportunities, define technical direction, and deliver products from concept through production.
  • Built a zero-to-one digital finance platform in Argentina, setting technical direction and leading delivery from initial concept and architecture through launch.
  • Delivered production platforms for more than five global organisations, adapting architectures and delivery approaches across customer environments to accelerate innovation and reduce time to market for AI-enabled products.

Engineering Tech Lead

McKinsey & Company, Digital Labs

2018 - 2020 · Berlin, DE

  • Led engineering for an AI-powered financial platform integrating open banking infrastructure with intelligent transaction analysis.
  • Designed scalable cloud-native architecture supporting secure aggregation of customer financial data across multiple banking providers.
  • Delivered Romania's first open-banking application, enabling customers to manage accounts across multiple banks through a single digital experience.

Head of Engineering | Interim CTO

Generation.org

2017 - 2018 · Mexico City, MX

  • Led global engineering strategy and platform development for Generation's international workforce development programmes.
  • Scaled technology platforms and engineering operations to support rapid international expansion while improving platform reliability and delivery velocity.
  • Enabled expansion from 4 to 16 countries, supporting a threefold increase in global programme delivery.
  • Established engineering capability for strategy, people, and delivery.

Snr. Software Engineer

McKinsey & Company, Digital Labs

2016 - 2017 · Johannesburg, ZA

  • Developed an AI-enabled customer matching platform connecting insurance customers with advisers using predictive recommendation models.
  • Designed scalable backend services and data pipelines supporting intelligent lead generation and customer routing.
  • Increased customer conversion by over 10% through deployment of machine learning-driven matching capabilities.

Engineer in Residence

JoziHub (AfriLabs)

2014 - 2016 · Johannesburg, ZA

  • Coached early-stage technology founders and engineering teams on software architecture, product engineering, and the path from prototype to scalable product.
  • Built prototypes and technical proofs of concepts across web and mobile platforms.
  • Supported founders on product engineering design decisions from zero to N.

Graduate Researcher & Engineering Lecturer

University of Johannesburg

2014 - 2016 · Johannesburg, ZA

  • Designed and delivered undergraduate software engineering courses covering algorithms, software architecture, and modern development practices.
  • Supervised and coached software engineering teams, helping students translate theory into practical, production-oriented delivery.
  • Contributed to machine learning research output in the Computer Science department

Software Engineer

Accenture

2013 - 2014 · Johannesburg, ZA

  • Developed enterprise advanced analytics solutions for global mining clients as part of Accenture's digital transformation practice.

Software Engineer

Standard Bank

2012 - 2013 · Johannesburg, ZA

  • Built enterprise banking software supporting core digital banking products.

Applied AI Expertise

LLM Applications and AI Agents

  • Production LLM Applications
  • Multi-Agent and Agentic Workflow Development
  • Tool-Using AI Systems (MCP, Function and Tool Calling)

Knowledge and Reasoning

  • Retrieval-Augmented Generation (RAG)
  • Semantic Search and Vector Retrieval
  • Enterprise Knowledge Systems and Vector-Store Pipelines

Model Development and Evaluation

  • Prompt Optimisation and AI Evaluation
  • Classical Machine Learning (Scikit-Learn)
  • Experimentation and Model Benchmarking

AI Platforms

  • AI Platform Architecture
  • Model Serving and Inference
  • Human-in-the-Loop AI Systems

Technical Expertise

Programming

Python • TypeScript • Java • Go

AI & Machine Learning

LangChain • LangGraph • MCP • PyTorch • Scikit-Learn

Infrastructure

Kubernetes • Docker • Terraform • Helm

Distributed Systems

Temporal • RabbitMQ

Cloud

Azure • AWS • GCP

Leadership

  • Engineering team leadership
  • Coaching and mentoring
  • Technical hiring and assessment
  • Customer-embedded delivery
  • Technical discovery and architecture
  • Executive stakeholder management
  • AI platform and product strategy
  • Production reliability
  • Cross-functional product and engineering leadership
  • Zero-to-one capability building

Education

MSc Computer Science (Machine Learning)

University of Johannesburg

Dissertation: BOLEPI: A Machine Learning Framework for Forecasting Project Outcomes

  • Designed and evaluated a neuroevolution-based forecasting framework that combined ANNs, evolutionary search, and ensemble learning to predict project outcomes.
  • Built the full prototype pipeline, including synthetic dataset generation, neural-network training, and repeated experimental validation with confusion-matrix metrics.
  • Diagnosed overfitting and improved generalisation via bootstrap aggregating, increasing test accuracy from 57% to 78%.

BSc Computer Science & Informatics

University of Johannesburg

Dissertation: Built an explainable AI market-forecasting prototype using Gene Expression Programming, giving financial analysts interpretable rule-based reasoning behind trade recommendations.