Parhat

Senior engineering · Applied AI · Ireland

Engineering judgement for serious software work.

I build production-grade Python, AI, computer-vision, and data systems, with the leadership range to move from architecture to delivery without losing sight of the business problem.

Arrange a call

For privacy, I share the full CV directly with relevant hiring teams and collaborators. If delivery fails, I follow up manually.

Evidence

A track record measured in production outcomes.

~$500M

annual savings from a patented computer-vision manufacturing system

80%

stoppage reduction at selected conveyor-line bottlenecks

30+

engineers interviewed, hired, and onboarded across multiple levels

15+

engineers mentored toward senior-level impact and ownership

Patents

Granted US patent

Systems and Methods for Monitoring Chain Health

US patent application

Systems and Methods for Identifying and Alerting Footprint Over-Cycle Risks on Multi-Product Assembly Lines

Positioning

What I bring

I work best where the problem is technical, operational, and commercial at once: enough ambiguity to need judgement, enough consequence to need disciplined execution. The right fit is work where senior engineering means more than implementation — shaping the path, earning stakeholder trust, and staying close enough to the details to make good decisions.

My background is production-heavy: computer vision, backend platforms, manufacturing analytics, optimisation, data acquisition, and applied AI — now extending into LLM and agentic workflows through focused independent systems.

Capabilities

Where I can help quickly

Production AI & Data Systems

  • Python services, data acquisition, ingestion pipelines, and analytics workflows
  • Computer vision and ML delivery with PyTorch, TensorFlow, YOLO, Scikit-learn, and Pandas
  • LLM and agentic workflow projects that retrieve, rank, reason over, and operationalise research or market data

Backend & Infrastructure

  • Microservices, distributed systems, REST APIs, Kubernetes, RabbitMQ, and CI/CD
  • Production diagnostics at the data layer, with attention to reliability, maintainability, and review discipline
  • Security-focused engineering background with OWASP-aware delivery

Leadership & Delivery

  • Technical strategy and roadmapping from ambiguous operational problems to shippable systems
  • Stakeholder alignment from engineering and industrial engineering through operations and senior leadership
  • Hiring, onboarding, mentoring, and recurring technical learning programmes for growing engineering teams

Experience

Career in brief

Aug 2017 to Feb 2026

Software Engineering Lead

General Motors

  • Architected and led a patented computer-vision system across data ingestion, ML pipelines, and production deployment, generating roughly $500M in annual savings.
  • Designed optimisation systems that reduced conveyor-line stoppages at critical bottlenecks by up to 80% and turned shop-floor pain points into engineering roadmaps.
  • Acted as the bridge between engineering, industrial engineering, operations, and senior leadership, translating business priorities into technical strategy.
  • Built engineering capability by hiring and onboarding 30+ engineers, mentoring 15+ engineers, and running technical learning sessions for 20-30 developers.

Mar 2026 to Apr 2026

Contract Senior Software Engineer

Fidelity Investments

  • Built Python-based data-acquisition systems using Playwright and Bright Data for quantitative research and sentiment-analysis pipelines.
  • Resolved production data-layer issues that were blocking downstream analytics and proposed improvements to code-review and delivery workflows.

Selected Projects

Independent Systems

ScholarFlow

An LLM-powered research assistant that retrieves, evaluates, and synthesises academic literature, surfacing research directions and key insights — growing into a full research lifecycle platform for postgraduate students.

Signal Lab

An algorithmic trading research system combining systematic scanning and filtering (SMA-20, momentum, rolling VWAP, MACD) with LLM-generated sentiment analysis and conviction scoring — live for crypto, paper-trading across equities.

Let's talk

For work that needs technical depth, clear ownership, and mature delivery.

A short call is enough to understand the problem, the team, the expectations, and whether the fit is real on both sides.

Arrange a callSend details