Full Stack Developer

As a Full-Stack AI Intern at Cognitive Videos, you'll work with our core dev team to build and optimize backend pipelines for AI-powered video generation. You'll get hands-on experience with media processing (FFmpeg, GStreamer), video AI model inference, agent frameworks, and graph databases on modern cloud infrastructure.

Backend Video Pipelines
The core responsibilities for the job include the following:

• Help design and implement backend pipelines for video generation and transformation.
• Integrate and orchestrate tools like FFmpeg and GStreamer for video encoding, decoding, compositing, and post-processing.
• Build services that chain together video AI models (open-source and third-party APIs) for tasks like generation, editing, and analysis.

Model Inference And AI Services

• Implement and optimize Python / PyTorch inference code for video and multimodal models.
• Contribute to batching, scheduling, and resource-efficient execution of model inference, including GPU-aware logic where applicable.
• Wrap models in robust APIs (e. g., FastAPI/Flask) that can be consumed by other services or frontends.

Agents, Tools, And Memory

• Work on agentic workflows using frameworks like LangGraph (and similar ecosystems).
• Implement and maintain agent tools for video processing, retrieval, and orchestration.
• Design and experiment with agent memory (short-term, long-term, episodic) using a combination of caches, vector stores, and graph databases.

Caching And Performance

• Implement caching layers for: Expensive model inferences, Frequently accessed assets (video segments, thumbnails), Agent context, and conversation state.
• Work with tools like Redis or similar technologies to reduce latency and cost.
• Contribute to performance profiling and basic optimizations (CPU/GPU usage, I/O, memory).

Graph Databases And Data Modeling

• Help design and implement graph database schemas for representing entities such as projects, scenes, clips, and workflows.
• Work with graph DBs (e. g., Neo4j / similar) to model relationships and power agent reasoning and retrieval.
• Write queries and small utilities to analyze and visualize graph data.

Cloud Infrastructure (GCP / AWS)

• Assist in deploying services to GCP and/or AWS, using managed services where appropriate.
• Work with: Compute (e. g., GCE/EC2 Cloud Run/Lambda), Storage (e. g., S3 GCS), Networking basics (load balancers, APIs, security groups/IAM).
• Contribute to simple infrastructure-as-code or configuration where needed.

Testing And Reliability

• Write unit, integration, and regression tests for: Python backend services, Model inference modules, Video processing pipelines.
• Help set up or extend CI/CD workflows (linting, tests, basic checks before deploy).
• Contribute to monitoring/logging hooks (e. g., logging inference times, error rates).

Requirements

• Currently pursuing a Bachelor's or Master's degree in Computer Science, Software Engineering, AI/ML, or a related field.

Core Programming

• Strong proficiency in Python (data structures, OOP, async basics).
• Comfortable working in a Linux development environment and using Git

AI / ML Fundamentals

• Solid understanding of machine learning / deep learning basics.
• Hands-on experience (coursework or projects) with PyTorch for training or inference.
• Ability to read and adapt open-source model code (e. g., diffusion/transformer models, vision models).

Backend Engineering

• Experience building basic backend APIs/services in Python (FastAPI, Flask, Django, or similar).
• Understanding of REST APIs, JSON, HTTP, and basic authentication concepts.

Media/video Processing

• Familiarity with FFmpeg (CLI usage, basic filters, transcoding) and/or GStreamer
• Understanding of core video concepts: codecs, containers, frame rate, resolution, and bitrates.

Cloud And DevOps Basics

• Exposure to at least one cloud platform (GCP or AWS can be via coursework or personal projects.
• Basic understanding of containers (Docker and how services are packaged and run.

Data And Storage

• Working knowledge of relational (SQL) and at least one NoSQL database.
• Interest in or exposure to graph databases and graph data modeling (nodes, edges, traversals).

Testing Mindset

• Familiarity with testing frameworks in Python (e. g., pytest, unittest).
• Ability to write simple unit tests and debug failing test cases.

Nice-to-Have (Big Plus)

• Prior project experience with:
• LangGraph, LangChain, or other agent frameworks.
• Building tool-using LLM agents (calling APIs/tools, RAG pipelines, memory).
• Graph DBs such as Neo4j, ArangoDB, TigerGraph, etc.

Experience With

• Redis or similar cache stores.
• Vector databases (e. g., Pinecone, Weaviate, Qdrant, Milvus) for retrieval/memory.
• GPU-accelerated workloads (CUDA basics, optimizing PyTorch inference).
• Contributions to open-source projects in video, ML, or agents.
• Portfolio of projects (GitHub, personal site) showcasing: Video/ML pipelines, Backend services, and Interesting AI demos.

Soft Skills

• Strong ownership and willingness to dive into unfamiliar stacks (video, agents, graph DBs).
• Curiosity about state-of-the-art video generation and AI agent systems.
• Ability to communicate clearly, document work, and collaborate with a small, fast-moving team.
• Comfortable iterating quickly, receiving feedback, and improving designs.

This job was posted by Nishanth Sirivolu from Cognitive Videos.