Skip to content
ASAbubakar SohailSystems / Product / AI
01Origin02Journey03Missions04Systems05Contact
Notes

Chapter 00Engineering the journey

Engineering the journey.

Senior Software Engineer building scalable products, intelligent systems, and reliable platforms.

I work across full-stack product development, backend architecture, production AI, cloud infrastructure, and developer experience—turning difficult technical problems into systems teams can confidently ship and operate.

Explore my workView experienceEngineering notesRésumé
Glasgow, United KingdomAvailable for selected opportunities
System model / interactiveOperational
01Engineeringas a connected system
Backend
Boundaries that hold

APIs, data models, and services shaped around real product responsibilities.

Scroll to begin

Chapter 01 / Career route

Built through ownership.

The route from end-to-end client delivery to leading production work across multi-tenant platforms, AI pipelines, and international product teams.

2020Now
032023 — presentSenior Software EngineerDevsinc·International product teamsImproved API response times by 20–30%, shipped production AI workflows, and strengthened reliability across application, data, and infrastructure layers.
The challenge

Turn complex, multi-tenant product requirements into systems that remain fast, observable, and supportable as they scale.

What I built

Backend services, GraphQL and REST APIs, AI-enabled analytics, FastAPI data pipelines, and cloud delivery workflows across marketplace and SaaS products.

Leadership signal

Led delivery with junior engineers, ran code reviews and debugging sessions, and helped new hires build confidence in project standards.

  • TypeScript
  • Node.js
  • FastAPI
  • Rails
  • GraphQL
  • PostgreSQL
  • AWS
  • Docker
022023Software EngineerTechnocares·Product engineeringReduced frontend loading time by 20% and improved the consistency of production issue resolution.
The challenge

Stabilize and evolve full-stack applications while improving a slow, fragmented user experience.

What I built

Frontend workflows, backend APIs, data flows, and production fixes across Node.js, React, Angular, and .NET systems.

Leadership signal

Worked across disciplines to reproduce issues, clarify root causes, and land maintainable fixes.

  • Node.js
  • React
  • Angular
  • C#
  • .NET
012020 — 2022Freelance Software EngineerIndependent·Remote client deliveryDelivered end-to-end products for international clients and established a practical foundation in product ownership.
The challenge

Translate incomplete client briefs into dependable products with clear scope, delivery, and support.

What I built

More than 20 full-stack applications covering authentication, integrations, data processing, deployment, and responsive product interfaces.

Leadership signal

Owned discovery, technical decisions, implementation, deployment, and client communication.

  • Node.js
  • React
  • Python
  • MongoDB
  • PostgreSQL

Chapter 02 / Selected missions

Systems with a reason to exist.

A small set of product stories: the constraint, the architecture, the decision, and the result. No technology theatre.

01
EM
AI & automation

Completed mission / AI & automation

Employer Reputation Navigator

An intelligence pipeline that converts fragmented employer and market data into decision-ready reputation signals.

Problem

Recruitment vendors needed usable insight from high-volume, unstructured data spread across reviews, CRM, billing, and market sources.

Open system notes
Architecture

FastAPI services on AWS Lambda orchestrate OpenAI and Gemini analysis before normalizing outputs into Snowflake for Looker dashboards.

Decision

Separated ingestion, model analysis, normalization, and presentation so every stage can be retried, inspected, and evolved independently.

Trade-off

Serverless execution reduced idle cost and operational weight, while requiring strict timeout, batching, and failure-handling boundaries.

ImpactTurned qualitative market noise into repeatable metrics and near-real-time vendor intelligence.

  • Python
  • FastAPI
  • AWS Lambda
  • OpenAI
  • Gemini
  • Snowflake
02
TR
Product & distributed systems

Completed mission / Product & distributed systems

TrueCar

A multi-tenant automotive marketplace serving real-time pricing and inventory through a distributed product platform.

Problem

Millions of buyers need accurate, local vehicle data without sacrificing search performance, SEO, or frontend responsiveness.

Open system notes
Architecture

Next.js and React at the experience layer, GraphQL resolver orchestration over Rails services, with PostgreSQL and Redis behind the domain.

Decision

Used server rendering for discoverability, resolver-level data shaping for efficiency, and targeted caching for high-read workflows.

Trade-off

Aggressive caching improves latency but raises invalidation risk, so caching remained explicit and domain-aware.

ImpactSupported a high-traffic buying journey with transparent pricing and faster access to local inventory.

  • Next.js
  • React
  • GraphQL
  • Ruby on Rails
  • PostgreSQL
  • Redis
03
TA
Platform engineering

Completed mission / Platform engineering

Talivity

A recruitment technology marketplace that unifies product discovery, CRM context, and campaign operations.

Problem

Recruitment teams were navigating disconnected tools, stale marketing data, and inconsistent campaign workflows.

Open system notes
Architecture

A TypeScript product surface backed by Node.js services, PostgreSQL, and integration adapters for HubSpot and Beehiiv.

Decision

Kept external integrations behind explicit adapters and synchronization jobs to prevent vendor-specific logic leaking into the core domain.

Trade-off

Near-real-time synchronization was more resilient than hard real-time coupling and easier for teams to operate.

ImpactCreated a coherent software discovery and campaign workflow across a fragmented vendor ecosystem.

  • TypeScript
  • Node.js
  • React
  • PostgreSQL
  • HubSpot
  • Beehiiv
04
VU
Commerce & product engineering

Completed mission / Commerce & product engineering

VU Custom

A product personalization system connecting customer creativity to fulfillment-ready production data.

Problem

Buyers needed confidence in highly customized products while production teams needed precise, automatable order specifications.

Open system notes
Architecture

Rails domain workflows and a React configuration experience coordinate 2D/3D previews, order data, and fulfillment handoff.

Decision

Made the preview part of the validation loop and treated production data—not the visual editor—as the source of truth.

Trade-off

Rich previews add complexity, so the workflow prioritized deterministic order output over purely visual flexibility.

ImpactReduced ambiguity between ecommerce orders and the production process for personalized products.

  • Ruby on Rails
  • React
  • PostgreSQL
  • 3D preview
  • Automation

Chapter 03 / Systems thinking

Architecture is a chain of promises.

Each layer has one job: protect the layer behind it from accidental complexity while keeping the whole path observable.

Reference flow / production productRequest → response → signal
  1. 01

    Experience

    Fast, accessible product surfaces that make system state understandable.

    • React
    • Next.js
    • TypeScript
  2. 02

    Contract

    Deliberate boundaries for validation, composition, and predictable change.

    • REST
    • GraphQL
    • Node.js
  3. 03

    Domain

    Services organized around product responsibilities—not framework convenience.

    • FastAPI
    • Django
    • Rails
  4. 04

    State

    Durable data models, clear ownership, and caching with explicit invalidation.

    • PostgreSQL
    • Redis
    • Kafka
  5. 05

    Operations

    Repeatable delivery with signals that make production behaviour debuggable.

    • AWS
    • Docker
    • Kubernetes
    • Terraform
    • CI/CD
    • Datadog
Cross-cutting signal

Prometheus · Grafana · Datadog · structured logs · traces · deployment health

How I use the stack

AI / RAG systems

Retrieval, evaluation, fallbacks, human-readable outputs, and cost are designed as product concerns—not added after the model works.

Async processing

Queues and events absorb variable work while idempotency, retries, and visibility keep failures safe to operate.

Cloud infrastructure

Containers, infrastructure as code, and CI/CD turn delivery into a repeatable system instead of a collection of heroic steps.

Chapter 04 / Selected impact

Better is measurable.

Performance work matters when it changes how a product feels, how a team operates, or what a business can reliably do next.

01

11s → 1.5s

Page-load recovery

Profiling, query reduction, and a tighter delivery path turned a blocking wait into a responsive product flow.
02

≈30%

Faster API responses

Resolver, service, and database-path optimisation improved critical request latency.
03

≈25%

Less over-fetching

Purpose-built GraphQL contracts reduced unnecessary data transfer and downstream work.
04

20+

Products delivered

End-to-end ownership across client discovery, architecture, implementation, deployment, and support.
Reliability is also an outcome

Test-driven delivery reduced avoidable UI regressions, while explicit async workflows and better observability made production issues faster to isolate and safer to recover.

Chapter 05 / Operating principles

The engineer behind the systems.

I approach engineering as a product discipline. The strongest solution is rarely the most elaborate one; it is the one that gives users a clear experience, gives teams a maintainable path forward, and gives operators enough signal to act when reality deviates from the diagram.

My work sits naturally between technical depth and product context. I can move from a database or distributed workflow to a customer-facing decision, explain the trade-off to technical and non-technical stakeholders, and keep momentum without leaving a fragile system behind.

I am especially interested in AI-enabled products, cloud platforms, developer experience, and systems that need to scale across teams and markets. From Glasgow, I work comfortably with international teams across the UK, Europe, Singapore, and timezone-friendly global environments.

  1. 01Make the system legible before making it clever.
  2. 02Optimise the path users and operators actually feel.
  3. 03Use AI where it creates leverage—and design for uncertainty.

Final chapter / Start a conversation

Have a difficult system or an ambitious product? Let's make it real.

Open to senior engineering and technical consulting conversations

I'm happy to speak with engineering leaders, founders, recruiters, and technical clients about senior product, backend, AI, platform, and consulting work.

abubakarsohail83@gmail.com
LinkedInProfessional profileGitHubCode & open workRésuméDownload PDF
Glasgow, UKInternational direction

ASDesigned & engineered by Abubakar Sohail

Engineering notesBack to origin