Ahmed Aslam.M

AI Engineer with 1+ year of experience building production-grade LLM systems, multi-agent architectures, and real-time AI pipelines. Skilled in OpenAI ecosystem, LangChain, and AWS. Built AFDE, a 4-agent autonomous financial decision engine with MCP-based live data and self-reflection. Focused on taking agentic AI systems from prototype to production.

ML Full Stack Developer Intern at HUM1AI (2024-12 – 2026-01)

Designed and delivered production-grade autonomous AI systems, LLM workflows, and multimodal RAG pipelines for enterprise automation and customer engagement solutions.

• Architected and deployed autonomous multi-agent systems using OpenAI Agent SDK for task automation, orchestration and AI decision-making.
• Implemented MCP servers to automate integration across enterprise and external systems, improving efficiency by ~30%.
• Built high-performance FastAPI/Flask services for model serving, document automation and workflow orchestration.
• Deployed scalable AI infra on AWS EC2, achieving ~25% latency reduction through system optimization.
• Developed RAG pipelines using LangChain and Pinecone for retrieval-based contextual reasoning across structured and unstructured data.
• Integrated OpenAI Assistants with secure file ingestion, chunking and metadata-based document routing.
• Implemented secure S3-based storage pipelines with user-level isolation for sensitive document workflows.
• Built Twilio-based AI voice support and multimodal conversational avatars with ElevenLabs TTS.
• Designed MLOps pipelines using AWS SageMaker and Postman CI for testing, deployment, rollout, and API monitoring.
• Implemented enterprise-grade OAuth authentication for secure system access.
• Conducted testing, logging, error monitoring, and system health checks for production reliability.
• Fine-tuned and integrated Transformer models using Hugging Face for NLP applications.

AI & ML Engineer Intern at Tamizh - TN Startup (2024-06 – Present)

Contributed to an AI-driven wearable 'smart specs' system by building CV models, OCR pipelines and gesture-recognition solutions optimized for real-time performance.

• Developed object detection models using Roboflow, Labelbox, and Florence fine-tuning.
• Implemented multilingual OCR (English, Tamil, Hindi, Telugu) using EasyOCR.
• Built a Mediapipe-based hand gesture recognition system enabling cursor control.
• Created GPU-accelerated client-server architecture for CV model deployment; conducted device-level performance tuning.
• Documented pipelines and results for internal knowledge transfer.

B.Tech in Artificial Intelligence And Data Science – B. S. Abdur Rahman Crescent Institute Of Science And Technology (2025-05)

HSC – Sunshine Senior Secondary School (2020-06 – 2021-03)

SSLC – Apex Pon Vidyashram (2018-06 – 2019-03)