# Mohammed Adnan

**Backend & AI Infrastructure Engineer** based in San Francisco, CA.

I build backend systems in TypeScript, Python and Go: APIs, event pipelines and LLM agent infrastructure, deployed on Kubernetes across AWS, GCP and Azure.

- Email: techadnank9@gmail.com
- GitHub: https://github.com/techadnank9
- LinkedIn: https://linkedin.com/in/techadnank9
- Résumé (PDF): https://techadnank99.web.app/resume.pdf
- Site: https://techadnank99.web.app

## About

I'm a backend engineer who owns the whole path from the database to the deploy. That's meant designing REST and gRPC APIs, moving high-volume event streams through Kafka, tuning query patterns until the latency actually drops, and containerizing services so releases stop being an event.

For the last year I've been building LLM and agent infrastructure full time: an execution engine with a C++ core and a Python node API, multi-agent systems with RAG and tool calling, and a voice agent running live in production. The evaluation discipline comes from training humanoid locomotion policies in simulation, where you don't get to claim something works without measuring it.

**Highlights:** 7 years in production · 40% ml data-retrieval latency cut · 30% deployment time cut

**Previously worked on:** Neudesic (IBM), Centific, Guidesly, Yoofoo Technologies

## Skills

- **Languages:** Python, C++, TypeScript, JavaScript, Go, Java, SQL
- **Backend & Data:** Node.js, Express, REST, gRPC, GraphQL, Kafka, Microservices, PostgreSQL, MongoDB, Neon, Supabase, Firebase, BigQuery
- **AI & Cloud:** CrewAI, RAG, Vector databases, Tool calling, Agent evaluation, PyTorch, TensorFlow, MuJoCo, Isaac Lab, AWS, GCP, Azure, Docker, Kubernetes, CI/CD

## Experience

### Independent AI Engineer & Builder, Independent (AI Infrastructure & Applied LLMs)
*May 2025 to Present*

- Designed and built an LLM pipeline engine with a C++ execution core and a Python node API, keeping the execution loop fast while extension nodes block, fail and retry without stalling the graph.
- Built multi-agent systems on CrewAI with prompt engineering, RAG and tool/function calling, including an adversarial review loop that scores its own confidence for downstream consumers.
- Deployed a production voice ordering agent with time-aware greetings, advance-order handling and dynamic variable injection, running live for a restaurant, and shipped the React and Next.js interfaces over it: streaming responses, tool-call visualization and agent traces.
- Trained humanoid locomotion policies with PPO in MuJoCo and Isaac Lab, building the simulation and evaluation tooling around them.
- Competed in AI and robotics hackathons including Physical AI Hack, Greptile at YC and MongoDB Build Lab, shipping prototypes in 24 to 48 hours.

### Team Lead, Centific
*May 2024 to May 2025*

- Led delivery of AI-powered enterprise modules in Python (Pandas, scikit-learn), architecting the data-processing automation and coordinating scope with product and QA.
- Built Node.js and Express microservices across AWS and Azure with Kubernetes to cut API response times, and deployed Python data pipelines on Kafka.

### Senior Software Engineer, Guidesly
*Sep 2023 to Apr 2024*

- Owned full-stack delivery of the booking platform: designed REST and gRPC APIs, optimized MongoDB query patterns to cut ML data-retrieval latency 40%, and built React Native layouts for iOS and Android.
- Led cloud modernization on AWS and GCP, containerizing microservices with Docker and Kubernetes to cut deployment time 30% and improve scalability.

### Full Stack Developer, Neudesic Technologies (IBM)
*Mar 2022 to Sep 2023*

- Built the Hours of Service module for the Mastery logistics platform, ingesting high-volume driver logs in real time through Kafka against federal compliance rules.
- Integrated custom ML models for driver-behavior anomaly detection, cut over-fetching through GraphQL query optimization, and used BigQuery and GCS for warehousing.

### Software Engineer, Yoofoo Technologies
*Mar 2019 to Mar 2022*

- Built GoZego property-management software (TypeScript, Firebase) automating tenant communications, payments and maintenance.
- Designed the Reverse Logistics platform in React Native for warehouse returns.

## Education

B.Tech, Computer Science & Engineering, TKR College of Engineering and Technology (2015 to 2019)

## Availability

Open to backend & AI infrastructure roles. If you're building infrastructure that has to hold up under real traffic, or agent systems that have to be more than a demo, I'd like to hear about it.