Hi, my name is
Humza Tareen.
I ship AI solutions with customers.
I sit at the intersection of engineering, enterprise delivery, and client success. I've fine-tuned LLMs for Apple, Meta, and Bytedance, delivered GenAI integrations for Fortune 500 clients across Mulesoft, Salesforce, and Kafka, and built agentic automations that cut client workloads by 60%. I ship faster using AI-native tools like Cursor, Codex, and Claude Code — including this very website. Currently providing services to Turing and actively open to new challenges.
About Me
I'm an AI engineer who thrives where technology meets customer outcomes. My career has taken me from training LLMs for the world's most demanding tech companies (Apple, Meta, Bytedance), to leading a globally-distributed team of 10+ across 3 continents, to delivering GenAI solutions directly to Fortune 500 enterprise clients across Mulesoft, Salesforce, and Kafka.
What sets me apart is that I don't just write code — I ship solutions that solve real client problems. I've scoped and delivered POCs for financial services clients, built AI agents that reduced customer workloads by 60%, architected integrations across enterprise iPaaS platforms, and hold IBM Cloud Sales and Mulesoft certifications that let me speak both the technical and business language. I also bring peer-reviewed research published at IEEE EMBC.
Client & Enterprise Delivery
AI & LLM Systems
Full-Stack Engineering
Where I've Shipped
Software Engineer @ Turing
- Architected and shipped 10 production microservices on GCP (Cloud Run, GKE, AlloyDB, Pub/Sub, Cloud Tasks) powering an AI agent evaluation platform — including the core evaluation engine, automated LLM scoring service, RAG retrieval pipeline, RL training arena, and authentication gateway.
- Diagnosed and resolved a critical AlloyDB SSL connection failure affecting 5% of evaluation tasks by tracing it to a connection pool lifecycle mismatch — reducing failure rate to 0.01%.
- Built an automated LLM scoring service with idempotent Cloud Tasks delivery, LiteLLM proxy integration for model-agnostic evaluation, and Redis-based real-time state tracking — handling thousands of evaluations daily with zero duplicates.
- Designed an event-driven pipeline using Pub/Sub (35 topics with DLQs), Cloud Tasks, and Cloud Scheduler to orchestrate 6-phase evaluation workflows with independent retry and zero silent failures.
- Conducted a systematic security audit across the platform, identifying critical vulnerabilities (SQL injection, hardcoded credentials, unauthenticated endpoints) and implementing fixes via GCP Secret Manager, parameterized queries, and structured logging.
Pod Lead @ Turing
- Promoted to lead a globally-distributed team of 10+ AI engineers across 3 continents, serving as the primary point of accountability for client delivery.
- Owned end-to-end execution of high-priority projects for key APAC clients, delivering model evaluations under accelerated timelines while managing client expectations and escalations.
- Built quality control frameworks and onboarding workflows that maintained consistently high output and client satisfaction in a fully remote, async environment.
LLM Python Developer @ Turing
- Developed and evaluated domain-specific LLMs for Apple, Meta, and Bytedance — directly contributing to the quality of models used by billions.
- Specialized in RLHF, advanced reasoning tasks, and Code Interpreter development to enhance model capabilities.
- Engineered complex evaluation suites to identify and resolve subtle model failures, improving final product accuracy and safety.
Software Engineer @ Royal Cyber Inc.
- Scoped, built, and delivered GenAI solutions directly for Fortune 500 enterprise clients using Mulesoft, Salesforce, and Apache Kafka — owning the full lifecycle from technical discovery to production deployment.
- Fine-tuned Llama models on AWS Bedrock to build domain-specific copilots for financial services clients, translating complex client requirements into high-accuracy AI products.
- Co-built 'RC AI OPS', a GenAI agent (Python, LangChain) deployed to client environments that automated error resolution and improved backend operational efficiency by 50%.
Technical Trainee @ Royal Cyber Inc.
- Built and deployed AI agents for enterprise Middleware Technologies (Mulesoft, IBM ACE, IBM MQ, ODM, Salesforce OMS), reducing client admin and developer workloads by 60%.
- Delivered AI agents accessible through Teams, Slack, and web — ensuring seamless adoption across customer organizations using Copilot Studio, Chainlit, LangChain, and Bot Framework.
Intern — Middleware @ Royal Cyber Inc.
- Earned Mulesoft Certified Developer (Level 1). Delivered multiple customer-facing POCs integrating Mule 4 APIs with Kafka, Flink, Slack, Gmail, Salesforce, and ServiceNow.
Deep Learning Researcher @ NCAI, Pakistan
- Authored a peer-reviewed paper published at IEEE EMBC 2025 on an award-winning explainable AI system for glaucoma detection (1st Prize, SEECS Open House).
- Achieved state-of-the-art results using attention-driven CNNs, Grad-CAM explainability, and a custom segmentation pipeline (Python, PyTorch, Detectron2).
Software Engineer @ InLights
- Led the Priority Signals team, owning the product roadmap and revamping core development and deployment processes.
- Directed the internship and MVP development program, translating client requirements into working prototypes that resulted in multiple industrial contract wins.
Mobile Application Developer @ InLights
- Engineered a mobile application prototype for optimizing emergency vehicle response times using Flutter & Firebase.
Software Engineer @ Bitnine Global Inc.
- Contributed to Apache AGE's core, developing methods for the 'agtype' datatype to enhance PostgreSQL graph database integration.
- Built an internal enterprise automation framework for cross-platform testing (Windows, CentOS, Red Hat, Debian, Ubuntu).
- Authored technical blogs driving the company's developer marketing and online presence.
Software Engineer Intern @ WALEE
Full-stack MEAN Stack development. Built scalable features, performed Twitter data analysis, led code reviews.
Core Contributor @ Unify
Open source contributor to Unify-Ivy. Implemented and tested scipy integration functions across multiple backends.
Mobile Developer @ Sparkleo Technologies
Built Flutter/BLE apps with Firebase. Developed sparkleo.io (Next.js), achieved first-page Google ranking via SEO, configured CI/CD pipeline.
Mobile App Developer @ Freelance
Built PalMy, a location-based social app (Flutter, REST APIs, Firebase realtime chat, Google APIs, multi-language support).
Additional Roles @ NUST & Others
Multiple mobile & application development roles at NUST (LMS app, guest navigation app). Junior developer at Shader Bytes. Campus Ambassador at SayabiDevs and Royal Cyber.
Technical Toolkit
Enterprise Integration & iPaaS
AI & LLM Systems
Backend & Infrastructure
GTM & Client Delivery
Mobile & Web
AI-Native Development
Tools & Practices
Education & Credentials
National University of Sciences & Technology (NUST)
Bachelor of Science, Computer Science
October 2020 — June 2024
IEEE EMBC 2025 Publication — OptiGuard
Featured Articles
Technical deep-dives and lessons from production
Loading articles...