Mayank Singh

The project involved developing the first Indian multiplayer battle royale game, Underworld Gang Wars. My role included creating all the game logic in C++ and maintaining the FPS for low-end devices.

Death Race is an adrenaline-fueled AAA-quality racing game unlike any other. It propels players into a fast-paced universe where the line between racing and combat blurs. Users compete against AI-controlled cars in a relentless quest for supremacy with diverse powers at their disposal. Players engage in intense vehicular warfare where they can take direct aim and shoot at rival cars or lock onto adversaries for precision strikes. Every race becomes a tactical showdown where skill and strategy reign supreme. The game features an AAA-quality immersion and is Windows and VR-compatible.

The main project is Underworld Gang Wars (UGW), which is India's first battle royal game. My part was to create all the game logic in C++, as we had to maintain the FPS for low-end devices. It is a multiplayer game.

A 3D shopping environment with mini-games.

One of the key projects involved the development of a 3D shopping environment that integrated interactive mini-games. This project aimed to provide users with a novel and engaging shopping experience, combining the practicality of online shopping with the enjoyment of immersive 3D interactions.

We embarked on an innovative venture to create Penhaligon's Metaverse, a groundbreaking 3D virtual museum that redefined how clients could showcase their products. This metaverse offered a unique and immersive way to experience and understand the products. It was a combination of an interactive product display, real-time production insights, engaging mini-games, and multiplayer interactions.

It is an open-world game in which players can roam and interact with different NPCs. NPCs will auto-spawn around the main player and can join the fight with the player. It is a GTA type of game.

FEATURES
• Players can pick up guns and grenades and do hand combat.
• Different types of AIs spawn throughout the level in runtime.
• Different AI will react to different situations, as they can equip guns.
• Runtime navigation generation for AIs.
• Aim offset and headshot implementation.
• AI hearing and sight perception
• AI can pick up guns and grenades and do hand combat.
• Petrol station blast. If a player or AI shoots at a petrol station, it will explode and apply damage to nearby actors.

It is a racing-type game in which players can use destructive weapons to destroy others' vehicles.

FEATURES
• Chaos vehicle with sports suspension and wheel alignment.
• Engine Sound dependency on RPM.
• Car backlight using Material Parameters collection.
• Smoke exhaust system.
• Skid system (tire scratch).
• Nitro System. Players will get limited nitrous once it is used. Nitrous will be auto-generated.
• Speedometer using Material only (Complex UI)
• Rear mirror system.
• Mini map using Material.

In this engaging project, I was part of a team that embarked on the creation of a 3D version of the classic Indian board game, Ludo. The primary objective was to bring the traditional game into the virtual realm, offering players a rich and interactive experience. The 3D multiplayer Ludo game allowed players to select their preferred colors, play animations for movement and piece captures, and feature captivating visual effects (VFX) to enhance gameplay.

In this exciting project, I was part of a dedicated team that developed a 3D multiplayer billiards game. Our goal was to bring the authentic billiards experience to the digital realm, enabling players to compete against each other, apply spin, and choose from various cue sticks with unique characteristics. The game was designed to offer a realistic and immersive experience, allowing players to host matches, join games, and even spectate other players' matches.

It is a Counterstrike-type game where you can compete with other players. You can send messages to teams or to your own teammates.

FEATURES
• Players can pick up guns and grenades and can do hand combat.
• Different types of AI spawn throughout the level in runtime.
• Different AI will react to different situations, as they can equip guns.
• Runtime navigation generation for AIs.
• Aim offset and headshot implementation.
• AI hearing and sight perception.
• AI can pick up guns and grenades and do hand combat.
• Everything is properly replicated in C++.

It was an n8n-based application designed to autonomously create 10–20 minute videos using agentic AI workflows. The user only needed to provide the name of a novel, after which the system intelligently generated a structured narrative by breaking the story into multiple coherent sequences. Each sequence was then converted into corresponding video visuals and audio narration using AI-powered generation tools.

The application automatically stitched all video and audio segments into a complete, polished video, ensuring narrative continuity and quality. In addition to content creation, the workflow also handled end-to-end automation, including publishing the final video directly to the target platform without any manual intervention.

Overall, the solution demonstrated a fully automated, scalable content production pipeline that leveraged agentic AI for story generation, media creation, assembly, and distribution.

Also wrote an application on top ot the TTS Diya model, which was totally in Python.