Pavan Kulkarni

CEO

Bengaluru, Karnataka, India10 yrs 7 mos experience

Key Highlights

AI leader with 11+ years in aerospace AI solutions.
Expert in LLM-powered systems and predictive maintenance.
Holder of 17+ invention disclosures and 4 patents.

Stackforce AI infers this person is a leader in Aerospace AI solutions with expertise in Machine Learning and AI Security.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Machine LearningAi SecurityData AnnotationResearch

Other Skills

Scikit-LearnTensorFlowAzure DatabricksLarge Language Models (LLM)Knowledge Graph-Based Natural Language ProcessingWord EmbeddingsLoRAPredictive AnalyticsAnomaly DetectionAmazon Web Services (AWS)MLflowAWS SageMakerMicrosoft AzureFastAPIAgile Methodologies

About

AI leader with 11+ years of experience driving enterprise AI solutions in the aerospace domain. Currently a Technical Manager at Pratt & Whitney, leading AI/ML initiatives across design, manufacturing, and MRO—delivering impact in predictive maintenance, quality assurance, and operational efficiency. I specialize in LLM-powered systems, including end-to-end RAG architectures and Agentic RAG frameworks for intelligent decision support and automation in complex engineering workflows. My work also leverages ontologies and knowledge graphs to enable structured reasoning, contextual retrieval, and explainable AI in safety-critical environments. Previously led AI security and responsible AI efforts at Bosch, focusing on securing ML and Generative AI systems against adversarial threats while improving model transparency and trust. Holder of 17+ Invention Disclousres (4 patents) and 7+ publications in NeurIPS, ICASSP, ICVS, and SecDev. Passionate about advancing AI-driven transformation in aerospace systems.

Experience

10 yrs 7 mos

Total Experience

Average Tenure

Current Experience

Pratt & whitney

Technical Lead (Data Science & Analytics)

Oct 2024 – Present · 1 yr 8 mos · Bengaluru, Karnataka, India · On-site

I lead a high-performing data science team delivering production-grade AI/ML solutions across the aerospace lifecycle—design, manufacturing, and MRO—enabling data-driven decision-making in safety-critical environments.
I also serve as the IP Focal for Pratt & Whitney DT India, driving innovation strategy, guiding invention disclosures, and aligning with RTX’s IP framework.
Key Contributions:
Engine Reliability & MRO Optimization:
Leading AI initiatives for engine health monitoring and advanced diagnostics (EHM/ADEM), including anomaly detection, RUL estimation, and condition-based maintenance, along with AI-driven root cause analysis using Physics Informed Neural Network (PINNs), Visual/Large Language Models (VLMs/LLMs).
Enterprise GenAI, LLMs & Agentic RAG:
Designing LLM-powered systems including end-to-end and Agentic RAG architectures for engineering knowledge, maintenance documentation, and decision support.
Enabling multi-step reasoning and workflow automation via agentic frameworks
Leveraging ontologies and knowledge graphs for semantic retrieval, contextual reasoning, and explainability.
Building scalable semantic search and maintenance analytics solutions on Azure to enhance root cause analysis and decision support.
Computer Vision & Multimodal AI:
Automating borescope inspections and defect detection (using U-Net, YOLO, DeepLabv3); applying multimodal AI for welding symbol interpretation and manufacturing quality checks (using RAGs, and GPT4.5)
Leadership & Delivery:
Driving AI strategy, leading cross-functional teams, managing stakeholders, and delivering end-to-end solutions from research to production. (frameworks: Agile/Scrum)
Innovation & IP
Leading invention disclosures and fostering a strong culture of innovation.

Scikit-LearnTensorFlowAzure DatabricksLarge Language Models (LLM)Knowledge Graph-Based Natural Language ProcessingWord Embeddings+11

Bosch global software technologies

Senior Research Engineer (AI/ML) - AIShield

Dec 2021 – Oct 2024 · 2 yrs 10 mos · Bangalore Urban, Karnataka, India

As a Senior Research Engineer, I have made significant contributions to AI security and Explainable AI across various domains, including Computer Vision, Tabular Data, Time Series, NLP, and LLMs. My key accomplishments include:
Conducting research on adversarial attacks against generative AI models and implementing defense mechanisms to enhance model robustness.
Developing advanced algorithms to detect and prevent model extraction attacks.
Conducting extensive research on evasion and poisoning attacks and improving detection and mitigation methods.
Building and testing adversarial robust AI models to reduce evasion attacks.
I have authored few significant publications (NeurIPS, ICVS, SecDev, ICDDS) and played a key role in filing 14+ invention reports in AI Security, potentially leading to patents.

Image ProcessingProject ManagementExploratory Data AnalysisAmazon Web Services (AWS)AI SecurityScikit-Learn+19

Technopro india

2 roles

Lead Machine Learning Engineer

Promoted

Apr 2021 – Dec 2021 · 8 mos

Core Areas of Expertise:
1. Precision Farming
Focus Areas: Machine Learning, Deep Neural Networks
2. Artificial Intelligence in Industry 4.0
Focus Areas: Machine Learning, Convolutional Neural Networks, Reinforcement Learning
3. Neural Machine Translation/NLP
Focus Areas: Deep Neural Networks, Text Analysis & Processing
4. Data Annotation & Labeling | Team Management | Scene Understanding
Led and managed a high-performing team of 18 data annotators on large-scale image and video labeling projects for Nexar, a leading ADAS (Advanced Driver-Assistance Systems) company. Oversaw end-to-end annotation workflows including quality assurance, workforce allocation, tool optimization, and deadline compliance.
Specialized in:
2D & 3D Bounding Boxes, Semantic Segmentation, Instance Segmentation, and Keypoint Annotation
Scene Understanding & Object Contextualization for autonomous driving
Sensor Fusion Labeling (e.g., LiDAR-Camera)
Annotation tool customization (CVAT, Labelbox, SuperAnnotate, V7, etc.)
Setting up annotation guidelines, reviewer pipelines, and quality metrics (IoU, mAP, QA-QC standards)
Proven track record in delivering high-quality, production-ready labeled datasets that significantly accelerated the model training pipelines for perception systems.

Autoregressive ModelsSamplingExploratory Data AnalysisNoSQLScikit-LearnTensorFlow+22

Senior Machine Learning Engineer

Aug 2020 – Apr 2021 · 8 mos

Core Areas I Have Worked On:
Precision Farming
Advanced Driver Assistance Systems (ADAS)
Healthcare AI
AI-based Smart Security Systems
Team Leadership:
Leading a team of 8 members in research, 9 members in product development and customer delivery.

Autoregressive ModelsSamplingExploratory Data AnalysisNoSQLScikit-LearnTensorFlow+19

Indian institute of science (iisc)

3 roles

Senior Project Associate

Nov 2017 – Aug 2020 · 2 yrs 9 mos

Areas I worked on:
1. Text-to-Speech Synthesis: Language/Linguistic Modeling, Text Analysis & Processing, Speech Processing, Deep Neural Networks, VOCODER(Synthesis).
Languages TTS built: Kannada, English(Indian), Hindi, and Telugu.
2. Epoch Extraction: Gammatone Filter Bank, Signal Processing, Speech Processing.
Publications:
1. Epoch Estimation from a Speech Signal Using Gammatone Wavelets in a Scattering Network, published in IEEE - ICASSP-2020.
Other activities:
I have also involved in writing two project proposals in medical image processing. One is on Glaucoma severity classification and the other one is Artificial Intelligence (AI) solution to Wireless Capsule Endoscopy (WCE). Project proposals are submitted to various central government schemes under the Department of Health Research, BIRAC-Grand challenge India, BIRAC-Biotechnology Ignition Grant(BIG)

Linear ProgrammingAutoregressive ModelsSamplingResource ManagementScikit-LearnTensorFlow+18

Senior Research Associate

Promoted

Aug 2016 – Aug 2020 · 4 yrs

Linear ProgrammingSamplingResource ManagementScikit-LearnTensorFlowFine Tuning+12

Project Assistant

Aug 2015 – Mar 2017 · 1 yr 7 mos

Projects I worked on during this tenure are:
1. Text-to-Speech (TTS) Synthesizer (DST funded project):
Areas Concentrated: Speech Processing, Signal Processing, Machine Learning, Natural Language Processing.
Topics: Text Tokenization, Text Transcription, Text processing using HTK Toolkit, Language Modelling(n-gram), Hidden Markov Models.
Languages TTS built: Kannada, Hindi, Tamil, Telugu, and Marathi. Bilingual Kannada-English, Hindi-English, and Telugu-English.
2. Epoch-based Time & Pitch Scaling - an Android App Development:
Areas concentrated: Speech processing, Signal Processing, Android Application Development. Link: https://arxiv.org/pdf/1801.06492.pdf
3. Voice Conversion - Pitch scaling from Source to Target
Areas Concentrated: Speech Processing, Signal Processing.