Kanak Dahake, MS

Amdocs (Amaiz AI)

Software Development Specialist - GenAI | June 2024 – Present | Pune, IN

• Developed key features for the GenAI platform, including a command step that generates structured commands for a pydantic data models from the user (human-in-loop) for seamless integration with external API systems.
• Engineered a context entity management system, utilizing Python, REST APIs, asyncio, Jinja2, and NetworkX to enable dynamic data fetching, dependency resolution, and integration with prompt templates, enhancing the platform's ability to handle complex, data model driven context-aware interactions.
• Designed and built a robust, configurable end-to-end Multimodal RAG pipeline architecture, integrating Docling, vector databases, and large language models to enable advanced retrieval and generation capabilities.
• Develop AI agents that enhance customer interactions by providing personalized billing support and pitch product recommendations in the telco sector.

PwC U.S. Advisory

Senior Associate - Innovation Hub | June 2022 – June 2024 | Mumbai, IN

• Spearheaded multiple production GenAI Enterprise applications, including Chat Assisted search (retrieval-augmented generation), Chat Insights (Statistics backed Agent on Structured Data), Supervised Content Creation, and other SME Chatbots, leveraging Azure Infrastructure like Cognitive Search, CosmoDB, App insights, etc.
• Led the development of InsightsGen, a GenAI accelerator, enabling data analysis with contextually relevant insights and visualizations using NLP, enhancing decision-making. This resulted in multiple client engagement wins for the firm.
• Filed 2 patents for designed algorithms.

Associate - Innovation Hub | Dec 2020 – June 2022 | Mumbai, IN

• Developed a Graph-based anomaly detection system for Azure IAM roles, resulting in a 20% reduction in unauthorized access incidents.
• Spearheaded the development of a Federated Machine Learning POC on a blockchain network, demonstrating the feasibility of secure and decentralized machine learning models.
• Designed and executed an NLU (Natural Language Understanding) Pipeline, streamlining information extraction from Clinical Trial documents and reducing manual processing time by 30%.
• Developed a novel approach to validate public addresses of multiple blockchain networks using cryptography such as EDDSA and ECDSA.

Digitate (Ignio AI), TCS

Technical Lead Cybersecurity | Jan 2020 – Nov 2020 | Pune, IN

• Developed Just-in-time access systems for IAM solutions using Python and MongoDB, improving access control and reducing unauthorized access by 35%.
• Implemented cloud security controls & Secure DevOps processes with code, automation, and data, improving the efficiency of security operations by 20%.
• Oversaw security for the SAAS delivery model, providing expert advice on secure architecture, design, and implementation.

Product Developer - Cognitive Threat Manager | Jun 2017 – Jan 2020 | Pune, IN

• Trained and deployed Deep learning-based Phishing URL and DGA domain detection models, reducing successful phishing attacks by 25%. Filled Patents for these systems.
• Developed ML services through a Python-based FLASK REST API system.
• Developed a Phishing campaigns detection system using Neo4J graph database and algorithms like label propagation and PageRank, leading to a 20% improvement in identifying malicious campaigns.
• Performed security data analysis, including UBA, providing static and dynamic alerting.