Software Engineer | GenAI Specialist | Full-Stack Developer
I build AI apps with reliable backends and frontends. Software engineer with experience in full-stack work, AI tools and setting up systems that run smoothly.
- 🔨 Engineer first - I focus on making products that work and fix real issues
- 🔭 My main focus is Diving deep into the ai native ecosystem 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, github.com/uzairakram6
Overview: Led development of key projects including a survey marketplace platform aggregating data from third-party providers and an automated invoice processing system using AI for efficiency gains.
Overview: Architected backend systems for property management operations and implemented testing strategies for pricing tools.
My primary focus for the past few days.
A powerful TypeScript implementation of an advanced multi-agent AI research orchestrator with Lead Agent coordination, specialized search subagents, automatic citations, and comprehensive logging capabilities.
Built entirely from scratch: Single-agent mode for quick queries, enhanced multi-agent mode for complex research, tool system integration, and CLI interface.
Technical implementation: Uses LangChain, OpenRouter for AI models, Redis for memory persistence, and tools like calculator, web search, and file operations.
Technical challenges solved:
- Implemented parallel processing for efficient agent execution
- Added adaptive configuration based on query complexity
- Integrated automatic citations and bibliography generation
- Enabled memory persistence and session management
- Developed extensible tool ecosystem with type safety
Tech stack: TypeScript, Node.js, LangChain, OpenRouter, Redis, DuckDuckGo search, and various npm packages like cheerio, axios, mathjs.
Quick note on this side project: Finance chat app that looks at your investment portfolio, suggests stocks with links to sources, and shares basic info on things like diversification and risk vs reward. Getting 200+ weekly users.
Another idea I'm working on: 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.
- Languages & DBs: TypeScript, GraphQL, PostgreSQL/MongoDB, Chroma DB, Mongoose
- Frameworks: Express.js, React/Next.js, Node/Nest.js, FastAPI
- Cloud & DevOps: Docker, Nginx, CI/CD, RabbitMQ, Digital Ocean, Microservices, Sentry, New Relic, Google Analytics, pytest, DeepDiff
- GenAI: LangChain, LangSmith, OpenAI, Hugging Face, Open Router, Gemini, AssemblyAI, Stable Diffusion, GPT-4 Vision, GPT-4o
- APIs & Security: Lucid Survey API, Trackier API, AES-256-GCM