~/motions
whoami
Karneeshkar V
>

Software Engineer building production-scale AI Inference and AI Engineering systems.
Currently Tech Lead at 2Cents Capital · Based in Bengaluru, India.

Contact me

My Projects

Valura AI — AI Investment Platform
React Native · Go · vLLM · Triton · Agno · LLMs · PostgreSQL · Supabase
2024 — Present · Live on iOS & Android
  • Shipped an AI-powered investment platform on iOS & Android from scratch to production — 4.8-star App Store rating, sub-500 ms cold-start latency, thousands of active users.
  • Led a team of five engineers building end-to-end LLM pipelines using both open-source and closed-source models, served at scale with vLLM and Triton Inference Server.
  • Designed scalable PostgreSQL architectures on Supabase Self-Hosted managing 10,000+ stock-market data points with efficient indexing, querying, and real-time updates.
  • Integrated computer-use automation and Agno-based agent workflows to generate financial insights at scale.
Level2 — Gym Management Platform
React Native · Go · AWS · PostgreSQL
2025 · Live on iOS & Android
  • Shipped a cross-platform gym management app live on the iOS App Store and Google Play — workout tracking, member management, and scheduling for gym owners and trainers.
  • Built an event-driven sync architecture achieving sub-second consistency across distributed mobile clients.
  • Built on-device facial biometrics — members enroll their biometrics from their phone and authenticate at the gym’s in-device terminal.
Naar — Restaurant Operations Platform
Next.js · Node.js · PostgreSQL · AWS
2025 · Live deployment
  • Built and deployed a full-stack restaurant platform — admin dashboard, kitchen display, and customer ordering — reducing order processing time by 50%.
  • Implemented role-based access control and a real-time order pipeline supporting concurrent multi-interface sessions.
naarkuwait.com
LLaMA 3.1 Fine-Tuning for Code Reasoning
Python · PyTorch · GRPO · Hugging Face · Gemma 3
Apr 2025
  • Fine-tuned LLaMA-3.1 8B using GRPO (DeepSeek-R1 methodology) with Gemma 3 as the reward model to improve chain-of-thought code reasoning.
  • Achieved 15% pass@1 improvement on custom benchmarks and 10% uplift on public coding benchmarks.
View on Hugging Face
Social Media Automation Engine
Python · FastAPI · Redis · AWS · Meta Graph API
2024
  • Built a cloud-native service processing 1,000+ daily interactions at 99.5% uptime with LLM-powered sentiment analysis (88% accuracy).
  • Implemented a Redis caching layer that reduced external API calls by 65% and kept p50 latency under 500 ms.
View on GitHub
VR Gesture-Recognition Gloves
ESP32 · Unity · C# · Python · Flex Sensors · IMU
Sept 2024
  • Built ML-powered VR gloves classifying 15+ hand gestures at 92% accuracy via flex-sensor signal processing on ESP32.
  • Created a configuration UI mapping gestures to 20+ system commands; integrated with Unity for gaming and assistive-tech use cases.
View on GitHub
Edge Sports Analytics System
ESP32 · PyTorch · OpenCV · Flask · MPU6050
2024
  • Deployed real-time computer vision on ESP32 for biomechanical analysis; fused IMU, VO2 max, and temperature sensor data.
  • Built a Flask/Plotly analytics dashboard and trained an ensemble classifier delivering actionable performance feedback to athletes.
View on GitHub
Smart Rhesus Macaque Deterrent System
Raspberry Pi · Jetson Nano · PyTorch · Computer Vision · Ultrasonics
Mar 2024
  • Developed an edge-deployed computer-vision system identifying macaques with 94% accuracy across varied lighting conditions.
  • Created a deterrent system using directional ultrasonic emitters, reducing campus intrusions by 90%.
  • Implemented power management reducing energy consumption by 35% for solar-powered deployment.
ProjectTheata — Backend
Go · PostgreSQL · REST APIs
2024
  • Open-source Go backend service exploring idiomatic Go patterns, structured logging, and clean modular architecture.
View on GitHub

My Experience

2Cents Capital
Software Engineer II / Tech Lead
Jul 2025 — Present · Bengaluru, India
  • First engineer hire; built and led a team of 10+ across mobile, AI, and quantitative finance — defined technical architecture, hiring bar, and engineering processes from day zero.
  • Shipped Valura AI (iOS + Android), an AI-powered investment platform, from scratch to production — 4.8-star App Store rating, sub-500 ms cold-start latency, thousands of active users.
  • Built a multi-agent AI system handling onboarding, support, and market intelligence — cut manual support tickets by 40% and reduced operational costs by 30%.
  • Designed algorithmic trading infrastructure: real-time market data ingestion, backtesting framework, and automated order execution pipeline processing live trades.
  • Established CI/CD, automated testing, and observability stack — reduced post-release bugs by 60% and maintained 95% on-time release cadence.
Visteon (Jaguar Land Rover)
Software Engineer
Jan 2025 — Jun 2025 · Chennai, India
  • Built diagnostic routines for JLR vehicle platforms with 96% anomaly detection accuracy; optimized SOME/IP middleware in C, cutting service-discovery latency by 25%.
  • Achieved 99.8% unit-test coverage across ASIL safety-critical code paths with zero compliance deviations.
Procter & Gamble
Automation Developer
Dec 2024 — Jan 2025 · Hyderabad, India
  • Deployed IIoT telemetry across 5 manufacturing lines ingesting 2M+ data points/day at 99.9% uptime — eliminated 40% of manual quality inspections.
Tynatech Ingenious
IoT Developer
May 2024 — Jun 2024 · Noida, India
  • Built REST API gateway connecting 50+ LoRaWAN edge devices to cloud; optimized UART layer for 30% throughput gain and implemented MQTT streaming at 100+ msg/s.
Hindustan Aeronautics Ltd (HAL)
R&D Developer
Aug 2023 — Sep 2023 · Bangalore, India
  • Optimized autopilot control algorithms reducing computational overhead by 22%; built a C++ simulation framework covering 200+ control-system test scenarios.
UpWork
Freelance Developer
Nov 2024 — Jan 2025
  • Built an enterprise-grade RAG system for legal applications reaching 92% retrieval accuracy with vector databases and fine-tuned LLMs.
  • Developed embedded audio-processing solutions on STM32/ESP32 platforms with 10% improved memory efficiency.
  • Delivered 5+ projects with a 4.9/5 average client rating and 100% satisfaction.
Intellect Design Arena
Software Developer
Sep 2024 — Dec 2024
  • Built PHP/WordPress financial-services websites with 10% faster page-load times.
  • Managed AWS infrastructure for 2+ web apps with auto-scaling at 99.9% availability; built real-time engagement and conversion dashboards.

Software Development Expertise

// what i build

Robust, scalable systems
from silicon to cloud.

I tackle every project with a strong foundation in software engineering principles — each solution reliable, maintainable, and efficient.

A systematic development approach combined with pragmatic problem solving drives delivery of high-quality, scalable applications.

My commitment is to ship software that is not only functional but optimized for performance and longevity.

Backend Development

REST APIs, FastAPI, Axum, Node.js, PHP

AI & Automation

LLMs, RAG, Function Calling, Workflow Automation

Cloud Solutions

AWS, DigitalOcean, Azure, Docker

Frontend Development

React, JavaScript, Responsive UI, UX Design

Tech Skills

Technologies I work with

Hands-on experience across the full stack — from embedded sensor firmware up through LLM fine-tuning and production cloud infrastructure.


  • Languages: Python, Go, TypeScript / JavaScript, Rust, C / C++, SQL, Bash
  • Backend & Cloud: Node.js, FastAPI, Flask, REST APIs, AWS (EC2, S3, Lambda), Docker, CI/CD, Git
  • Frontend & Mobile: React Native, Next.js, React, iOS / Android deployment
  • AI / ML: PyTorch, TensorFlow, LLM fine-tuning (RLHF / GRPO), RAG pipelines, multi-agent systems, OpenCV
  • Data: PostgreSQL, MongoDB, Redis, MySQL, SQLite, real-time streaming pipelines
  • Embedded: ESP32, STM32, Zephyr RTOS, LoRaWAN, MQTT, sensor fusion
Python 90%
Go 85%
TypeScript 80%
Rust 70%

About Me

Karneeshkar V - Photo
Who am I?

I’m a software engineer working across AI, fintech, and embedded systems. Currently Tech Lead at 2Cents Capital, where I was the first engineering hire and now lead a team of 10+ across mobile, AI, and quantitative finance.

I’ve shipped production iOS/Android apps with thousands of active users, built multi-agent AI systems running real customer journeys, and designed algorithmic trading infrastructure for live markets. Previously: Visteon (Jaguar Land Rover), Procter & Gamble, and HAL.

B.Tech in Electronics & Computer Engineering from VIT Chennai (8.66 / 10). Outside of work I’m a vim user (try pressing ? on this page), neovim config tinkerer, and into Linux daily-drivers.

Download Resume

Contact Me

Get in Touch

I'm always open to discussing new projects, creative ideas or opportunities to be part of your visions.

Name
Karneeshkar V
Email
karneeshkar68@gmail.com
Phone
+91 99109 77193
Based in
Bengaluru, India