# Hrishikesh Jadhav

> AI engineer in Germany with 4+ years shipping production AI and backend systems. I build LLM extraction pipelines, RAG and MCP systems, and the evaluation gates that decide what ships.

AI Engineer, Germany. https://www.hrishikeshjadhav.com

## Summary

I'm Hrishi, and at Toku I built the LLM extraction pipeline, RAG assistant and MCP server that run in daily production inside a global payroll platform. Before that I was a data scientist at GfK / NIQ in Nuremberg, building LLM classification, semantic retrieval and prediction models and running them in production with Python, AWS, Docker and GitLab CI/CD.

## Availability

Open to AI engineering roles in Germany from October 2026. EU Blue Card holder, authorised to work in Germany.

## Contact

- Email: knowhrishi.de@gmail.com
- Website: https://www.hrishikeshjadhav.com
- CV: https://www.hrishikeshjadhav.com/cv
- GitHub: https://github.com/knowhrishi
- LinkedIn: https://www.linkedin.com/in/knowhrishi

## Roles this profile fits

- [AI Engineer](https://www.hrishikeshjadhav.com): AI engineer in Germany with 4+ years shipping production AI and backend systems. I build LLM extraction pipelines, RAG and MCP systems, and the evaluation gates that decide what ships.
- [LLM Engineer](https://www.hrishikeshjadhav.com/llm-engineer): LLM engineer in Germany. I build extraction, classification and retrieval systems on self-hosted and hosted models, with evaluation gates that keep them correct in production.
- [Forward Deployed Engineer](https://www.hrishikeshjadhav.com/forward-deployed-engineer): Forward deployed engineer in Germany. I put AI tools into daily use with the teams that rely on them, and own deployment, documentation, onboarding and hands-on support.
- [Applied AI Engineer](https://www.hrishikeshjadhav.com/applied-ai-engineer): Applied AI engineer in Germany. I turn models into production systems: LLM pipelines with evaluation gates at Toku, prediction and retrieval models at GfK / NIQ.
- [AI Solutions Engineer](https://www.hrishikeshjadhav.com/ai-solutions-engineer): AI solutions engineer in Germany. I build AI tools and automations for the teams and customers who use them, and support them after launch.
- [GenAI Engineer](https://www.hrishikeshjadhav.com/genai-engineer): GenAI engineer in Germany. I build RAG assistants, LLM extraction and classification, and the evaluation that keeps them correct in production.

## Experience

### AI Engineer, Toku (10/2025-09/2026)

employed via WorkCo Germany GmbH, Frankfurt / Remote

Global employer-of-record and payroll infrastructure platform operating across 46 countries.

#### Internal AI tooling

- Shipped AI tooling into daily production across product, engineering and sales teams: an MCP server exposing internal systems to LLM clients, a RAG assistant (FastAPI, LangChain, pgvector), and n8n automations for onboarding, payslip processing and support triage. Owned deployment, documentation, onboarding and hands-on user support.
- Moved security review to pull-request time across thousands of pull requests by integrating automated LLM-assisted security review into GitHub CI for Toku's financial platform.

#### LLM pipelines and evaluation

- Cut payroll data intake from hours of manual work per cycle to minutes across production records by building a self-hosted LLM extraction pipeline (Python, FastAPI, PostgreSQL, DigitalOcean) that masks personal data before inference. (Self-hosted so that no employee data leaves controlled infrastructure.)
- Kept field-level extraction accuracy above the agreed threshold across every extracted field, checked against a large golden evaluation set, by building the evaluation harness and a CI regression gate that blocks any deploy below it. (So a change that lowers extraction accuracy cannot reach payroll data.)
- Brought LLM inference costs down to a fraction of the estimated cost of equivalent third-party API usage by deploying open-weight models on self-hosted GPU infrastructure.

#### Customer-facing platforms

- Built status.toku.com, Toku's public status platform: production components probed from the Cloudflare edge on a short fixed interval, automated incident detection and resolution, dependency rollups, edge-vs-origin failure isolation, and a JSON API and RSS feed. New incidents notify the team on Slack, and failures that need a code fix are picked up by a coding agent. (Probing the edge and the origin separately localises a failure to the CDN or the application before anyone looks at it.)
- Built Toku's customer-facing help centre and AI assistant (toku.com/help, Next.js) on Toku's Notion knowledge base, covering payroll, benefits, policy and country-guide articles across dozens of countries, with general-help and visa-support content in separate retrieval corpora and multi-model fallback across providers. (Separate corpora so immigration and work-authorisation questions never resolve against general payroll content.)

### Data Scientist, GfK - An NIQ Company (08/2022-09/2025)

Nuremberg

Earlier title: Working Student, Data Science (08/2022-07/2024)

- Reduced manual effort in product-taxonomy classification by around 30% by building an LLM classification service on Amazon Bedrock using prompt engineering and RegEx-based post-processing.
- Reduced catalogue lookup time by around 40% versus keyword search by building vector-embedding semantic retrieval over internal product catalogues.
- Improved simultaneous-viewer prediction accuracy by around 20% over the incumbent model by training and deploying a CatBoost model using 30+ engineered features from sociodemographics, temporal patterns and programme metadata.
- Kept production scoring pipelines running 24/7 without manual intervention by orchestrating S3 ingestion, feature generation, model scoring and automated integration tests in GitLab CI/CD, containerised with Docker on Linux.
- Built a TV-audience data-fusion pipeline using K-Nearest Neighbors and a genetic algorithm to match and clone panel households against large-scale return-path data.

### Software Developer (Internship), Sapio Analytics Pvt. Ltd. (04/2020-08/2020)

Mumbai

- Built COVID-19 decision-support models in Python (SEIRD, scikit-learn, SciPy) and forecasting dashboards (Plotly, AWS) used to inform Government of India lockdown and testing-strategy decisions, with a 4.8% RMSE reduction against the prior baseline.

## Education

- M.Sc. Artificial Intelligence Engineering (coursework and thesis), University of Passau, 2021-2024
- Honours Degree in Entrepreneurship, University of Passau, 04/2022-10/2022
- B.E. Computer Science, Savitribai Phule Pune University, Pune, 2018-2021

## Skills

- LLM systems and agents: LLM application design, RAG, knowledge retrieval, agentic AI, tool calling, MCP servers, structured outputs, prompt engineering, guardrails, LLM evaluation and regression testing
- Models and frameworks: LangChain, LangGraph, CrewAI, Claude SDK, OpenAI SDK, Amazon Bedrock, OpenRouter, Hugging Face, open-weight models, Whisper
- Backend and integration: Python, TypeScript, Node.js, FastAPI, Flask, Django, Next.js, React, REST API design, API integration, webhooks, OAuth, n8n, workflow orchestration
- Data and retrieval: PostgreSQL, Prisma, pgvector, Qdrant, FAISS, Pinecone, vector search, embeddings, data modeling, data pipelines
- Cloud and operations: AWS (Bedrock, S3), GCP, DigitalOcean (self-hosted GPU inference), Cloudflare, Docker, Linux, Git, GitLab CI/CD, GitHub Actions, Terraform, Kubernetes, observability, production monitoring, incident tooling
- ML and data science: PyTorch, scikit-learn, CatBoost, pandas, NumPy, feature engineering, model evaluation
- AI coding tools: Claude Code, Cursor

## Languages

English (C1), German (B1), Hindi (native), Marathi (native)

## Publications

### Ontology Evolution in Invasion Biology Using Large Language Models: A Hybrid Approach (2025)

Hrishikesh Jadhav, Tina Heger, Birgitta König-Ries, Alsayed Algergawy. LLM-TEXT2KG 2025, CEUR Workshop Proceedings Vol. 4020, pp. 195-206.

A hybrid pipeline that combines GPT-4 prompting and zero-shot extraction with classical ontology engineering to build and evolve INBIO, a core ontology for invasion biology, with domain experts validating new classes.

- PDF: https://ceur-ws.org/Vol-4020/Paper_ID_13.pdf
- Proceedings: https://ceur-ws.org/Vol-4020/
- INBIO ontology: https://bioportal.bioontology.org/ontologies/INBIO

### A Deep Learning Mobile Application based Sign Language Recognition for Aphasic Person (2021)

Hrishikesh Jadhav, Pushkar Dounde, Akash Pawar, Abhishek Muthange. Journal of Emerging Technologies and Innovative Research (JETIR).

An Android app that recognises sign-language gestures using Histogram of Oriented Gradients features with CNN and multiclass SVM classifiers.

- Paper: https://www.jetir.org/view?paper=JETIR2105465
- PDF: https://www.jetir.org/papers/JETIR2105465.pdf

## Projects

- Internal RAG learning assistant: NIQ/GfK HACKFEST 2024, Top 3. Built and demoed a retrieval-augmented learning assistant over internal knowledge within 24 hours, with document retrieval and audio ingestion via Whisper (LangChain, vector database, GPT, FastAPI).

## Awards

- 2024, NIQ/GfK HACKFEST: Top 3, with a RAG learning assistant built and demoed in 24 hours
- 2024, BMW Innovation Challenge: Selected participant, 24-hour challenge at BMW iFactory Dingolfing (DocCheck use case)
- 2020, IEEE Machine Learning Hackathon: 1st place
- 2020, HackCovid-19: Winner among 130 teams
- 2017, Smart India Hackathon: Winner (Ministry of Defence)

## German version

https://www.hrishikeshjadhav.com/de/llms-full.txt