name: Jayesh Manani
location: Stuttgart, Germany 🇩🇪 (originally from India 🇮🇳)
education:
degree: M.Sc. Computer Science - Autonomous Systems ✅
school: University of Stuttgart
current_status: Open to new opportunities 👋
background:
- AI & Data Engineer @ Mercedes-Benz AG (2 yrs · Production AI & Pipelines)
- 42 Heilbronn student (peer-to-peer, project-based coding school)
focus_areas:
- Autonomous AI Agents & LLM Systems (pydantic-ai · Vertex AI / Gemini · RAG)
- Cloud Engineering & Production MLOps (AWS · Azure · GCP · Databricks · Docker)
- Cybersecurity Engineering & SecOps (SIEM · SOC Triage · SC-200 · Threat Intel)
- Low-Level Systems & Concurrency in C/C++ (@ 42 Heilbronn)
superpower: >
Turning complex AI, security, and cloud data challenges into production-ready
autonomous systems that reduce human fatigue and scale reliably.
certifications:
- Microsoft SC-200: Security Operations Analyst ✅
- Generative AI with LLMs (DeepLearning.AI) ✅
- Oracle Cloud Infrastructure Generative AI Professional ✅
- AWS Machine Learning · Google Cloud Engineering · IBM Data Science ✅
open_to:
- AI Engineer / LLM Engineer
- Cloud / MLOps / Platform Engineer
- Cybersecurity / SecOps Engineer
- Autonomous Systems Engineer| 🤖 AI & LLM Systems | ☁️ Cloud, Data & MLOps | 🛡️ Cybersecurity & SecOps |
|---|---|---|
Autonomous Agents: pydantic-ai, Gemini, Vertex AI, Langfuse observability |
Multi-Cloud: Azure (Mercedes-Benz 2 yrs), GCP, AWS Machine Learning Specialty | SOC Automation: Real-time SIEM alert triage, VirusTotal & AbuseIPDB enrichment |
| Constrained Decoding: Custom logit masking & deterministic function calling | Production MLOps: MLflow, DVC, Databricks, Docker, CI/CD pipeline automation | Security Operations: Microsoft SC-200 certified, threat intelligence, log correlation |
| Enterprise RAG & Search: ChromaDB, CLIP multimodal search, zero-hallucination agents | Backend & Systems: High-concurrency FastAPI microservices, C/C++ (@ 42 Heilbronn) | Privacy & Compliance: Automated PII redaction pipeline (Presidio, spaCy, GDPR) |
An interactive AI representative speaking for my skills, experience, and background in real-time
An interactive, LLM-grounded AI Career Representative that allows recruiters, engineering managers, and collaborators to interview an AI twin trained on my verified background. It answers deep-dive technical questions, explores past experience (Mercedes-Benz, 42 Heilbronn, M.Sc. Autonomous Systems), highlights core engineering strengths and growth areas, and serves candidate data with zero hallucination.
Designed with a sleek, responsive dark UI, pre-built interview prompts (technical achievements, strengths, role fit), and instant PDF resume access.
LLM Grounding · Interactive AI Agent · FastAPI / API Integration · Vercel · Full-Stack · Candidate Screening
Autonomous security analyst assistant for SOC alert investigation
An autonomous AI assistant that analyses, prioritises, and investigates security alerts — the kind of work that burns out SOC analysts. It enriches events in real-time with VirusTotal and AbuseIPDB threat context, scores risk automatically, and supports natural-language conversational queries so analysts can ask questions instead of digging through dashboards.
Built with Gemini / Vertex AI via pydantic-ai, served over FastAPI, and evaluated with Langfuse for production-grade observability.
Building interactive agentic applications and automation tools — including my deployed AI Career Representative (Repo), an LLM-grounded candidate screening agent. Exploring how Gen AI and RAG architectures can be integrated into production workflows to eliminate repetitive manual queries.
Generative AI · LLMs · Python · Scikit-learn · NLTK · Transformers
Building autonomous agents that tackle real SOC workload - alert triage, threat enrichment, risk scoring. Combining my SC-200 security background with LLM agentic frameworks to ship tools that actually reduce analyst fatigue in production environments.
Generative AI · LLMs · Python · SC-200 · TryHackMe · Log Analysis
Engineering reproducible ML pipelines and scalable serving infrastructure — combining MLflow experiment tracking, DVC data versioning, container orchestration, and multi-cloud environments (Azure, GCP, AWS). Bridging the gap between experimental AI prototypes and resilient, enterprise-grade production systems.
MLOps · MLflow · DVC · Docker · Azure · Databricks · CI/CD · FastAPI
Deep-diving into low-level C from scratch - no extra libraries, only in-built libraries, no shortcuts. Peer-reviewing code every day, shipping projects like custom shells, libft, ft_printf, push_swap, GetNextLine, maze generators, Codexion (Classis dining philosopher problem), RAG, CallMeMaybe (Constrained Generation), all peer-reviewed. The hardest and most rewarding programming environment I've been in.
C · Python · C++ · Unix/Linux · Makefile · Algorithms & Data Structures
| # | Project | What It Does | Stack |
|---|---|---|---|
| 💼 | AI Career Representative (Live Demo) ⭐ | Interactive AI resume assistant & candidate screening representative grounded in verified candidate data | LLMs · Grounded Agents · Vercel · Full-Stack |
| 🛡️ | Autonomous SIEM AI Agent (Live Demo) ⭐ | AI agent for SOC L1 triage - alert enrichment (VirusTotal, AbuseIPDB), risk scoring, conversational queries | Gemini · pydantic-ai · FastAPI · Langfuse |
| 🧠 | Call Me Maybe | Constrained LLM decoding engine for 100% schema-compliant JSON function calling via logit masking | LLMs · Constrained Decoding · Logit Masking · Python · Qwen |
| 🚀 | MLOps Production Track | End-to-end production ML system (appliedAI Institute) with experiment tracking, DVC, CI/CD & orchestration | MLOps · MLflow · DVC · Docker · CI/CD · Python |
| 🔒 | STACKIT Data Vault | Privacy-preserving PII redaction pipeline for traffic logs using Presidio, spaCy NER & application-layer encryption | PySpark · Presidio · spaCy · Cryptography · GDPR |
| 🔍 | Multimodal Product Matcher | End-to-end multimodal search & matching system using CLIP embeddings, ChromaDB vector store, and MongoDB | CLIP · ChromaDB · FastAPI · MongoDB · Vector Search |
| ⚡ | Codexion Concurrency Engine | Multi-threaded simulation in C orchestrating POSIX threads, deadlock avoidance & FIFO/EDF scheduling | C · POSIX Threads · Mutexes · Real-Time Scheduling (EDF) |
| 🔄 | Push Swap Algorithm | Optimized 2-stack sorting algorithm in C engineered for minimum operation count and algorithmic complexity | C · Algorithms · Data Structures · Complexity |
👋 Currently open to new roles in GenAI, AI, Cybersecurity, Autonomous Systems, or impactful software engineering.
2025 ─ Student @ 42 Heilbronn (🇩🇪) ─▶ Present
└─ Peer-to-peer · No lectures · Project-based · Learning by shipping
Apr 2023 ─ Working Student / Master Thesis Student @ Mercedes-Benz AG (🇩🇪) ─▶ Mar 2025 ─▶ Completed ✅
└─ GenAI · RAG · AI/ML · Data ETL · Databricks · Llama3.1
└─ Python · Cloud CI/CD · Data Pipelines · MS Azure
└─ Real production environment at a global automotive engineering company
└─ Stuttgart, Germany 🇩🇪
May 2024 ── HiWi (Research Assistant) @ University of Stuttgart (🇩🇪) ─▶ Dec 2024
└─ Academic research support · University of Stuttgart
Apr 2022 ─ M.Sc. Computer Science @ University of Stuttgart (🇩🇪) ─▶ Mar 2025 ─▶ Completed ✅
└─ Autonomous Systems (Major)
Oct 2020 ─ Software Developer @ Sameeksha Capital (FinTech) (🇮🇳) ─▶ Mar 2022
└─ Built in-house trade processing platform from scratch
└─ GCP infrastructure · Workflow automation · Data pipelines
Jul 2019 ─ Research Associate @ IIM Ahmedabad (🇮🇳) ─▶ Oct 2020
└─ Research tooling for Prof. Samrat Gupta & Prof. Sanket Mohapatra
└─ Social network analysis · Community detection · Django/Flask · Research tooling
Jan 2019 ─ AI/ML Engineer @ Allevents.in (🇮🇳) ─▶ Jul 2019
└─ Duplicate listing detection · User clustering by interest
└─ Multi-label hierarchical text classification
└─ Future event prediction for Marketing
| Certification | Issuer | |
|---|---|---|
| 🛡️ | Security Operations Analyst — SC-200 | Microsoft |
| 🤖 | Generative AI with Large Language Models | DeepLearning.AI |
| ☁️ | Oracle Cloud Infrastructure Generative AI Professional | Oracle |
| 🌩️ | AWS Machine Learning Specialty | Amazon |
| 🌐 | Google Cloud Professional Data Engineer | |
| 📊 | IBM Data Science Professional Certificate | IBM / Coursera |
| 🔐 | Ethical Hacking Nanodegree | Udacity |
I write guides on Python, ML, and data engineering - from beginner to production:
- 💡 Data Driven Investor - data, AI, and tech perspectives
Got a data problem worth solving, or working on something interesting in AI/ML, Security or Autonomous Systems?
I'm always up for a good conversation.



