AI / ML ENGINEER
SAMPARK BHOL
Founding ML Engineer · Elsevier-published researcher. I build production-grade intelligent systems end to end — NLP, CV, predictive analytics, and quantum ML.
PLAYER 1
CHARACTER SELECT
Aspiring AI & Machine Learning engineer driven by an obsessive curiosity for building production-grade intelligent systems end to end — across NLP, computer vision, predictive analytics, and quantum machine learning. Currently a Founding ML Engineer at CLR3 Ventures, architecting autonomous LLM agent pipelines, transformer similarity engines, and full ML observability stacks from the ground up. Published Elsevier researcher and active open-source contributor who reads papers cover to cover, fine-tunes frontier models for fun, and ships code that lands in production.
VIT Vellore
2022 — 2026B.Tech, Computer Science & Engineering
Specialization in Blockchain Technologies
“Knowledge forged in stone, sharpened on silicon.”
- Distributed ledgers, smart contracts & consensus
- DSA · Operating Systems · DBMS · Networks
- Where the campaign began.
SEALED · 入学
THE LLM SUMMONER
Generative AI
Binds frontier models to his will — RAG, agents, and fine-tunes that actually ship.
Signature Spells
- RAG Pipelines
- LangChain / LangGraph
- LoRA / QLoRA Fine-Tuning
WPN · Staff of Retrieval
Career stats
Read the paper. Cover to cover.
Fine-tuned a frontier model. For fun.
Shipped it to production. It held.
CAMPAIGN
STAGES CLEARED
World 1: every role is a level conquered. Follow the path.
- LVL30
Founding Machine Learning Engineer (Intern)
@ CLR3 Ventures (Spizen Technologies / NoLimitNodes)Jan 2026 — Present◈ Bengaluru, India
Building the founding ML stack — agents, recommenders, and observability — from zero.
- Profitability prediction with ensemble gradient boosting on streaming Kafka pipelines — 92% revenue-forecast accuracy powering dynamic pricing
- Transformer embedding similarity engine (ANN on ClickHouse) + neural collaborative-filtering recommender at 10K+ qps in production
- Autonomous long-form content & synthetic-data pipeline orchestrating multi-stage LLM agents (plan → draft → summarize → validate)
- Founding ML infra on Kafka with Grafana drift/latency observability + CI/CD with A/B testing for safe rollouts
LOADOUT:KafkaClickHouseLLM AgentsGrafanaGradient Boosting - LVL26
Quantum Computing & Research Intern
@ QuantumX (Appetite Studio)Dec 2025 — Jan 2026◈ Remote
Hybrid quantum-classical ML + quantum-safe security research.
- Hybrid quantum-classical models — variational quantum circuits + classical NNs for combinatorial optimization
- Quantum optimization with QAOA & VQE; investigated post-quantum cryptographic architectures for AI model integrity
- Authored open-source Quark (quantum-circuit semantic-similarity embeddings + QASM dedup CLI); first-authored a journal under review
LOADOUT:QAOAVQEVariational CircuitsPythonQASM - LVL20
AI & ML Engineer Intern
@ CSM TechnologiesMay 2025 — Jul 2025◈ Odisha, India
Production RAG + transformer NLP, fine-tuned and shipped.
- Production RAG pipelines (LangChain, LangFlow, ChromaDB, Pinecone) — +25% domain accuracy via tuned chunking & prompt engineering
- BERT/GPT transformer apps for classification, NER, summarization with real-time inference + vector DB integration
- Fine-tuned LLMs with LoRA/QLoRA — 60% less training compute; Dockerized + Kubernetes multi-service deploys
LOADOUT:LangChainRAGLoRA/QLoRADockerKubernetes - LVL14
Deep Learning & Software Development Intern
@ National Informatics Centre, MeitY, Govt. of IndiaMay 2024 — Jul 2024◈ Odisha, India
Anomaly detection & predictive modeling on 1M+ government records.
- Anomaly detection (Isolation Forest, DBSCAN, Z-score, graph analysis) on 1M+ records — 95% precision-recall
- Predictive models (LogReg, SVM, Gradient Boosting, K-Means, PCA) with full EDA + feature engineering
- Automated ETL + warehousing in Python/SQL — 70% faster processing; dashboards adopted across departments
LOADOUT:Scikit-learnPythonSQLETLMatplotlib ↑ START • 4 STAGES CLEARED • NOW ↑
BOSS FIGHTS
PROJECTS
Select your opponent. Hover to land a hit and watch the boss HP drain.
CAIRN
Autonomous LLM experiment runner
▸ Boss staggered! Combo x2
An autonomous experiment runner letting LLM coding agents iteratively improve a codebase while keeping a persistent wiki of every attempt. Python daemon over msgpack RPC with SQLite WAL state, git-worktree isolation, hard budget caps, and an anti-Goodhart holdout evaluator — covered by 67+ tests.
BROWSER-LLM
Offline AI chatbot in the browser
▸ Boss staggered! Combo x3
Fully client-side AI chatbot with WebGPU hardware acceleration — sub-second token latency, zero backend, complete on-device privacy. Runs quantized GGUF & ONNX models via lazy weight streaming, KV-cache management, and memory-efficient attention on consumer hardware.
DATAWIZ-PRO
AI-powered AutoML analytics platform
▸ Boss staggered! Combo x4
AutoML analytics platform with Bayesian hyperparameter tuning, automated feature/model selection across regression, classification, and clustering. End-to-end CSV/JSON/SQL ingestion with automated preprocessing, outlier detection, and interactive BI dashboards for non-technical stakeholders.
LEAFGUARD AI
Client-side plant-disease diagnosis
▸ Boss staggered! Combo x5
Plant-disease diagnosis web app with a React frontend and TensorFlow.js for client-side CNN inference — real-time agricultural AI assistance with no server round-trip.
ASSIGNOFAST
Offline-first student assignment app
▸ Boss staggered! Combo x6
Smart assignment-management app for students — syncs assignments and timetables from Firebase while supporting full offline access via SQLite.
POWER-UPS
ITEM SHOP
npm-installable power-ups. Downloads = ammo. Tap the command to grab it.
cli-legend
v1.0.0A fun command-line roguelike dungeon game — ASCII graphics, turn-based combat, inventory.
gamedev-simple-utils
v1.0.0A simple utility package for game development.
dev-toolkit-sampark
v1.0.0A Gen-Z-inspired CLI toolkit for modern developers — from linting gently to shipping fast.
ARCHIVES
SCROLLS OF LORE
Ancient tomes of research, recovered from the vaults — each one unrolls as you descend.
Codex Vol. MMXXIV — 2 bound scrolls
Peer-reviewed relicsScroll 02 · Lex Arcana
An Energy-Efficient Hybrid Communication Protocol for Large-Area Wireless Sensor Networks
“A novel energy-efficient hybrid routing protocol for wireless sensor networks, delivering significant efficiency improvements for large-scale IoT deployments through optimized cluster-head selection and data-aggregation strategies.”
Scroll 01 · Lex Arcana
Hybrid Classical and Quantum Query Complexity for Hypothesis Verification Under Structured Priors and Bounded Phase Noise
“First-authored research on hybrid classical–quantum query complexity for hypothesis verification under structured priors and bounded phase noise — developed during the QuantumX internship alongside the open-source Quark library. Currently under review for publication.”
TRANSMISSIONS
LATEST ISSUES
Dispatches from the field — broadcasting live.
NeuroSymbolic AI: A View
Integrating neural networks with symbolic reasoning — can hybrid models overcome the limits of purely data-driven or rule-based AI?
Quantum-Enhanced NLP: The Future of Cloud AI
How quantum computing could revolutionize NLP tasks like sentiment analysis and translation in the cloud.
CERN in October: A Month of Milestones and Marvels
A recap of significant achievements, experiments, and groundbreaking discoveries at CERN during a pivotal month.
Quantum-Driven Gaming: AI Meets the Future
Exploring the intersection of quantum computing, AI, and the future of gaming at the cusp of a new tech era.
SIDE QUESTS
CO-OP MISSIONS
Pinned to the guild quest-board — take one on.
AI/ML Lead
2023 — 2025@ IEEE Computer Society, VIT
Led the AI/ML wing — events, workshops, and community projects.
Rewards
- Organized hackathons with 200+ participants
- Ran DL / NLP / CV workshops for students
Member
2023 — 2024@ Association of Energy Engineers
Active member contributing to energy-efficiency and sustainability initiatives.
Rewards
- Energy-efficiency research community
- Sustainability-focused collaboration
INVENTORY
SKILL BINDER
The full collection — every card on the page. Hover to foil, click to inspect.
Languages
9 CARDSAI / ML & Deep Learning
12 CARDSNLP / GenAI
11 CARDSData Science & Analytics
12 CARDSMLOps & Infrastructure
11 CARDSDatabases & Cloud
12 CARDSWeb & Tools
13 CARDSTROPHIES
ACHIEVEMENTS UNLOCKED
第01話 — Oracle Generative AI Cloud Infrastructure Professional
CONTINUE?
ENTER YOUR NAME
Available for work and collaboration. Got a quest? Insert coin and say hello.
FIND ME