$ cat ./profile/summary.md

Richard Hallett

AI-Augmented Software Engineer

I build production systems where humans orchestrate AI agents rather than write every line.

$ suggested queries:
>
$ ./fit-assessment --interactive

am I a good fit?

$ paste job_description.txt

honest assessment including where I'm not the best fit

$ cat ./summary.txt

summary

Software Engineer operating at the frontier of AI-augmented development. Building production systems where humans orchestrate AI agents rather than write every line. 4 years enterprise experience (React/TypeScript, Python), combined with 15 years as a Cognitive Behavioural Therapist—expertise in structured reasoning, communication, and human-AI collaboration patterns. Currently shipping multi-agent workflows daily using Claude Code as my primary development methodology.

$ ls -la ./skills/

skills matrix

$ cat ./skills/agentic.txt

[agentic] - differentiator

  • -Agent Orchestration (Claude Code)(Daily practice shipping multi-agent workflows)
  • -Context Engineering(Designing what information agents have at each decision point)
  • -Prompt Architecture(Reusable prompt patterns for consistent agent behaviour)
  • -File-based State Management(External persistence for agent memory beyond context windows)
  • -Agentic Loop Design(Autonomous task completion with validation gates)
  • -LLM API Integration(Claude, OpenAI API integration with streaming)
$ cat ./skills/core.txt

[core] - foundation

  • -React(4+ years, including enterprise scale at Brandwatch)
  • -TypeScript(Primary language for frontend and Node.js)
  • -Python(Automation, data processing, backend services)
  • -Node.js(API development, tooling, serverless functions)
  • -Next.js(App Router, Server Components, API routes)
  • -SQL(PostgreSQL, SQLite, query optimization)
  • -Git(Version control, branching strategies, CI/CD)
  • -Docker(Containerization for development and deployment)
$ cat ./skills/moderate.txt

[moderate] - proficient

  • -Vue.js(Production experience, prefer React)
  • -Angular(Project experience at Telesoft)
  • -Django(Collaborated with Django backends)
  • -AWS/Cloud(Vercel primary, AWS exposure)
$ cat ./skills/gaps.txt

[gaps] - not for me

  • -Legacy Java/C#(Not my focus area—greenfield AI development preferred)
  • -On-premise Infrastructure(Cloud-native approach, limited on-prem experience)
  • -Large Team Management(Mentored individuals, haven't managed teams of 10+)
  • -Mobile Development(Web-focused, no native iOS/Android)
$ ls -la ./experience/

professional experience

click [view context] to see the unflattened story behind each role

$ cat ./experience/oceanheart/readme.md

Founder & AI Systems Engineer

Oceanheart.ai(AI-augmented development studio and personal platform)

[featured]Apr 2024 – Present
  • -Built multi-agent orchestration systems using Claude Code, including automated job application pipelines with parser generation, match analysis, and document export
  • -Designed context-engineered workflows solving the 'context rot' problem through external persistence via file-based state, PRDs, and git commits
  • -Implemented agentic loops for autonomous task completion with acceptance criteria validation
  • -Created reusable prompt architectures for consistent agent behaviour across domains
  • -Shipped client websites (Next.js, Astro) with custom CMS and AI-powered features
Claude CodeMulti-agent orchestrationNext.jsAstroPythonFile-based state managementKernel MCP
$ cat ./experience/edited/readme.md

Software Engineer

EDITED(AI-driven retail analytics SaaS platform)

Nov 2023 – Mar 2024
  • -Developed React/TypeScript features for AI-driven retail analytics SaaS platform
  • -Built data visualisation components enabling clients to interpret ML insights
  • -Partnered with backend engineers (Python/Django) to optimize API integrations
ReactTypeScriptPythonDjangoData visualization
$ cat ./experience/brandwatch/readme.md

Software Engineer

Brandwatch(Enterprise social media analytics platform)

Jun 2021 – Nov 2023
  • -Contributed to enterprise data visualisation platform (Monitor project) serving millions of users
  • -Built scalable React components while modernising legacy Backbone codebase
  • -Mentored junior developers and facilitated knowledge-sharing sessions
ReactTypeScriptBackbone.jsEnterprise scale
$ cat ./experience/telesoft/readme.md

Full Stack Engineer

Telesoft(Cybersecurity applications)

Sep 2020 – Feb 2021
  • -Delivered secure features for cybersecurity applications (TypeScript, Angular, Node.js)
  • -Built Python utilities for data processing and automation
TypeScriptAngularNode.jsPythonSecurity
$ cat ./experience/sbs/readme.md

Software Developer

School Business Services Ltd(Financial management software for schools)

Mar 2019 – Mar 2020
  • -Created Vue.js frontend features for school financial management software
  • -Applied psychology expertise to simplify complex workflows and improve usability
Vue.jsJavaScriptUX design
$ cat ./background/psychology.md

psychology background

Cognitive Behavioural Therapist

15 years

15 years as a Cognitive Behavioural Therapist. This isn't just "soft skills"—it's structured problem decomposition, clear communication under uncertainty, and understanding how humans collaborate with systems.

Directly applicable to prompt engineering, agent behaviour design, and building AI tools that actually help people. CBT is fundamentally about changing thought patterns through structured intervention—which is exactly what prompt engineering does for AI.

# engineering applications
-Structured Problem Decomposition: CBT breaks complex problems into manageable components with clear intervention points. I apply this same approach to system architecture and agent workflow design.
-Communication Under Uncertainty: Therapy involves navigating ambiguous situations with limited information. This transfers to working with stakeholders, debugging unclear issues, and designing for edge cases.
-Prompt Engineering as Cognitive Restructuring: CBT changes thought patterns through careful questioning and reframing. Prompt engineering is the same skill applied to AI—guiding the model's 'thinking' through structured input.
-Human-Centered Design: Understanding how people actually think (not how they should think) is core to both therapy and UX. I design AI interfaces that match human mental models.
-Rubric-Driven Assessment: Clinical assessment uses structured rubrics to evaluate progress. I apply this to AI output validation—clear criteria, consistent measurement, actionable feedback.
$ cat ./education.txt

education

PGDip Cognitive Behavioural Therapy

Royal Holloway, University of London

2015

PGCert Cognitive Behavioural Therapy

University of Central Lancashire

2013

BSc Psychology

University of the West of England

2008
ongoing:

AI/ML systems, context engineering patterns, agentic workflow design

$ ./contact --collaborate

let's collaborate

looking for an AI systems engineer who combines technical excellence with human-centered design?

[get in touch][view portfolio]