Software Engineer | GenAI Specialist | Full-Stack Developer

I Despised AI for like 6 months because of how bad of an experince i had with gpt 3.5 but since sonnet 3.5 i have been bullish on AI and anything remotely related to AI , so if any of your projects contain "AI" in it thats all i need to get hooked

• 🔨 Engineer first - I focus on making products that work and fix real issues
• 🔭 My main focus is diving deep into the AI-native ecosystem and building deep research workflows
• 💼 Currently at Techverx working on backends and AI features
• 📝 Connect with me on LinkedIn: linkedin.com/in/uzair-akram
• 📫 Reach me at uzairakrum@gmail.com
• § GitHub: github.com/Uzair-Akram, My office account => github.com/uzairakram6

Key Projects:

• SumHuman AI (AI Video Generation Platform): Architected a multi-agent, Human-in-the-Loop (HITL) system to automate video generation from text prompts. Orchestrated a pipeline from research to final video synthesis using specialized agents (Research, Image Gathering, Fact-Checking) and integrated RunwayML for generation.
• GTS (Invoice Processing): Delivered a full-stack pipeline using FastAPI and GPT-4 Vision, reducing client workload by 120 hours weekly. Implemented automated correction loops with GPT-4o and Gemini, improving accuracy to 94%.
• Survey Marketplace: Led the MVP delivery for a survey aggregator, integrating Lucid Survey API and building an OfferWall module with fraud detection.

Overview: Architected backend systems for property management operations (Buildium) supporting 5,000+ units. Implemented RabbitMQ for asynchronous processing and fine-tuned MongoDB aggregation pipelines for a 2.5x speed increase.

An advanced AI research orchestrator with a Lead Agent that coordinates specialized subagents for automatic citations and comprehensive logging.

• Tech: TypeScript, OpenRouter, Redis (Memory Persistence).
• Features: Extensible Tool System (calculator, search, file operations) and support for multiple AI models.

Finance chat app that reviews investment portfolios, suggests stocks, and provides sourced financial info. Attracted 200+ weekly users and handles 350+ weekly queries.

• Tech: Next.js, Postgres, ChromaDB (Embeddings), Gemini.

Contributed to an open-source framework for building and deploying AI assistants, focusing on agents capable of planning and executing complex tasks.

AI platform that turns podcast or video recordings into show notes, with transcription, key points, timestamps, and easy publishing to places like YouTube or WordPress.

• GenAI & Agentic Systems: LangChain, LangSmith, CrewAI, LlamaIndex, RAG, Multi-Agent Systems, Agentic Design Patterns
• Generative Models: OpenAI (GPT-4 Vision, GPT-4o), Google Gemini, OpenRouter, RunwayML, AssemblyAI, Stable Diffusion
• Languages & DBs: TypeScript, Python, GraphQL, PostgreSQL, MongoDB, Mongoose, ChromaDB, Pinecone, Supabase (pgvector)
• Frameworks: FastAPI, Express.js, React, Next.js, Node.js
• Cloud & DevOps: Docker, Nginx, CI/CD, RabbitMQ, Microservices, Sentry, New Relic, pytest, DeepDiff
• APIs & Security: Lucid Survey API, Trackier API, AES-256-GCM