Jack Shiels

Co-Founder

Kuala Lumpur, Federal Territory, Malaysia15 yrs 11 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Founder of innovative AI startup shiels.ai.
Reduced energy wastage by over 80% through AI research.
Expert in LLM, NLP, and reinforcement learning systems.

Stackforce AI infers this person is a SaaS and AI solutions architect with a focus on enterprise and IoT applications.

Contact

Skills

Core Skills

Applied AiLlm SystemsReinforcement LearningFull-stack Development

Other Skills

Research and Development (R&D)Large Language ModelsC#LLMLarge Language Models (LLM)Start-up LeadershipASP.NETStartup DevelopmentStart-up VenturesArtificial Intelligence (AI)Generative AINatural Language Processing (NLP)AI product developmentOntology-driven designData extraction

About

Applied AI engineer and founder with experience building and deploying LLM, NLP, and reinforcement-learning systems across research and commercial settings. Founder of shiels.ai, where I have led the development of AI products including an RL-based research impact forecasting agent and an ontology-driven enterprise requirements and agent orchestration platform. Previously conducted AI research at UCL on trustworthy LLM systems for IoT environments, contributing to a model ecosystem that reduced energy wastage by over 80%. Earlier experience includes full-stack software engineering at Trayport on large-scale production systems in the Microsoft stack, and enterprise solutions work at Accenture’s REPL Group. Core strengths include Python, C#, LLM systems, applied research, full-stack engineering, and translating ambiguous business problems into working technical systems. Core Skills: Applied AI: LLM systems, NLP, reinforcement learning, prompt and workflow design, fine-tuning, evaluation, agent orchestration Languages: Python, C#, JavaScript, TypeScript, SQL Frameworks and Platforms: PyTorch, ASP.NET Core, React, SQL Server, Azure, AWS Engineering: full-stack development, API design, testing strategy, production deployment, cloud systems Research and Product: applied AI research, experimentation, technical writing, stakeholder discovery, product scoping

Experience

15 yrs 11 mos

Total Experience

2 yrs 9 mos

Average Tenure

2 yrs

Current Experience

Ucl

AI Software Engineer and Researcher

Jun 2024 – Mar 2025 · 9 mos · London Area, United Kingdom

Conducted applied research into adapting large language models to time-series and sensor-driven environments, using Python and PyTorch.
Contributed to the development of TRUST2, a trustworthy AI model ecosystem for IoT building control that paired black-box forecasting models with LLM-based decision explanation.
Helped build a live agent pipeline that combined anomaly forecasting, occupancy prediction, contextual prompting, and LLM-based HVAC control reasoning.
Supported evaluation of multiple LLMs for logical accuracy, truthfulness, and explainability in real-world control scenarios.
Contributed to a system that reduced energy wastage by over 80% while balancing energy efficiency and occupant comfort.
Worked across Linux and macOS environments, optimising models for live deployment on both CUDA and MPS hardware.
Reported technical findings and research outcomes to project stakeholders.

Large Language ModelsLLMGenerative AIApplied AI

Shiels.ai

Founder

May 2024 – Present · 2 yrs · Hybrid

Founded an applied AI company focused on LLM, NLP, and agentic systems for research and enterprise use cases.
Built ImpactAgent, a Prime Intellect reinforcement-learning agent based on Qwen, designed to forecast the likely impact of novel dementia research.
Led development of Flewnt, an ontology-driven data extraction and agent orchestration platform for enterprise requirements analysis and software delivery.
Designed Flewnt as an end-to-end workflow spanning stakeholder interviewing, document ingestion, ontology derivation, ambiguity detection, project-plan generation, DAG-based implementation planning, and guarded agent orchestration.
Developed supporting pipelines for custom model training, data extraction, LLM fine-tuning, and deployment across cloud and local environments.
Delivered custom GPT solutions.
Owned product strategy, technical architecture, client delivery, business development, and startup operations.

Research and Development (R&D)Large Language ModelsC#LLMLarge Language Models (LLM)Start-up Leadership+8

Trayport

Software Developer

Aug 2022 – Apr 2024 · 1 yr 8 mos · London, England, United Kingdom

Worked as a mid-level software developer on large-scale production codebases comprising several million lines, contributing features across enterprise trading software systems.
Delivered full-stack functionality across C#, ASP.NET Core, JavaScript, TypeScript, Razor, React, and SQL Server environments.
Took work from product discussion through implementation, testing, and production deployment.
Contributed to multilayer application architecture, authorisation flows, front-end components, automated testing, and production support.
Collaborated closely with engineers and product stakeholders using pair programming and shared delivery ownership.

C#ASP.NETFull-stack development

Repl group, part of accenture

Technology Delivery Analyst

Jan 2022 – Aug 2022 · 7 mos · London, United Kingdom

I worked as a solutions architect for a large-scale retail operations solution, being developed for a multinational firm. My responsibilities included documentation of technical requirements and solutions design for cloud and other systems.
I was accredited with AWS Cloud Practitioner and Azure Fundamentals certifications.

University of cape town

Information Systems Tutor

Apr 2021 – Jan 2022 · 9 mos · Cape Town, Western Cape, South Africa

Tutoring services for Information Systems students at the University of Cape Town. Modules: INF1002F/S and INF3003W.

Self employed

Full-stack Developer

May 2019 – Jan 2022 · 2 yrs 8 mos · Cape Town Area, South Africa

Designed and implemented two major information systems for clients while completing university studies.
Built and deployed a C# / ASP.NET learning management system on Azure for youth entrepreneurship education.
Built a dynamic medical notation system with support for custom forms, deployed within a major South African hospital.

Self-employed

Producer

Oct 2008 – Oct 2017 · 9 yrs · Cape Town Area, South Africa

As a teenager, I oversaw the creation of several video game titles on a broad range of technical platforms, including room-scale virtual reality. Some of our titles were awarded in international competitions.