SOFTWARE ENGINEER
Suhas R G
Building backend systems, developer tools, and hardware integrations.
PROJECT - Scout
AI-powered career roadmap generator. Given a goal and skill level, produces a personalised learning path with resource recommendations and project modules. Tradeoff: chose Gemini Flash over GPT-4 to stay within free-tier rate limits while keeping latency under 3s; added mock mode so the frontend develops independently of the AI backend.
STACK: Python · FastAPI · React · Google Gemini · Supabase · Upstash Redis
Read more →PROJECT - Shawty
High-performance URL shortener with custom aliases, link expiration, click analytics, and QR code generation. Redis caches hot redirects so the database is hit only on cache miss; async workers update click stats without blocking the redirect path, keeping p99 latency under 50ms.
STACK: Next.js · TypeScript · Supabase · Drizzle ORM · Upstash Redis
Link → GitHubPROJECT - Connect (Pi Player)
Remote media player for Raspberry Pi deployments. Exposes a REST API for playlist management; a daemon handles multi-format playback, looping, and auto-restart via systemd. Checksum verification prevents corrupted media from reaching the player. Built at Conekt-Labs to cut on-site maintenance visits by 60%.
STACK: Python · FastAPI · VLC · Upstash Redis · systemd · Raspberry Pi
Link → GitHubPROJECT - Drive Downloader
CLI tool for bulk-downloading files and folders from Google Drive. Supports both OAuth2 (interactive) and service-account auth (headless automation). Written in Go for a single static binary with no runtime dependency - chosen over Python because distribution to non-devs ruled out managing a venv.
STACK: Go · Google Drive API · OAuth2
Link → GitHubPROJECT - RAGX
Production-grade RAG service covering the full pipeline: document ingestion, chunking, embedding generation, vector indexing, retrieval, and answer synthesis. Evaluation harness tracks relevance and faithfulness scores. Streaming responses reduce perceived latency; rate limiting and caching control API cost.
STACK: Python · FastAPI · LangChain · OpenAI · Vector DB · PyTorch
Link → GitHubPROJECT - KTranslate
Telegram bot for real-time group-chat translation. Async message handling keeps the bot responsive under concurrent load; multiple translation engines are wired in with intelligent fallback so a single provider outage does not break the service. Live on Telegram at @k_nexus_bot.
STACK: Python · FastAPI · asyncio · Telegram Bot API · Docker
Link → GitHubCONEKT-LABS / IOT INTERN
Built an MVP Raspberry Pi media player system with 4G LTE connectivity for remote digital signage. Developed backend integration for remote playlist synchronisation and device health monitoring, cutting manual on-site maintenance by 60%.
OCT 2025 – PRESENT · BENGALURU · Python · FastAPI · Redis · systemd
FSMK.ORG / AI/ML INTERN
Designed and deployed an end-to-end RAG service - ingestion, chunking, embeddings, vector indexing, retrieval, and synthesis. Instrumented evaluation for relevance and faithfulness. Traded higher model quality for streaming and caching to hit latency and cost targets for a free-tier deployment.
AUG – NOV 2025 · REMOTE · Python · LangChain · FastAPI · OpenAI
VYAS LDM / FOUNDING TEAM MEMBER
Designed user file-management endpoints. Developed gRPC APIs in Go for microservices communication. Integrated backend with React frontend. Reduced deployment time 30–40% via CI/CD pipelines. Built a distributed Git service using Docker. Collaborated across a cross-functional team of 10.
MAY 2025 – PRESENT · BENGALURU / REMOTE · Go · React · Next.js · Docker · gRPC
I build software across the stack - from REST APIs running on Raspberry Pis to RAG pipelines serving AI products to full-stack web applications. My engineering instinct is to ship working systems first, then instrument and optimise where the data shows it matters. I care about production concerns: latency budgets, cost ceilings, failure modes, and the operational burden a system places on its maintainers. Currently interning at Conekt-Labs (IoT), with prior experience in AI/ML and full-stack development. B.E. Computer Science, HKBK College of Engineering (CGPA 8.66, 2026).