LinkedIn https://www.linkedin.com/in/dharanibabu2025

Models @ HuggingFace https://huggingface.co/aifinitydigital

Computational Mathematics in AI for Drug Discovery 

https://www.linkedin.com/pulse/two-proteins-one-conversation-how-numbers-predict-binding-babu-rss6c/

Cloud Contributions

https://medium.com/@dharanibabu_68045/a-free-video-conferencing-solution-on-amazon-cloud-b90f607326db

PROJECTS 

EpiDock AI: An antibody - antigen Interaction & Structural Prediction Platform

• EpiDock AI is an enterprise-grade structural biology pipeline designed for a top-tier pharmaceutical client to accelerate the discovery of targeted bio-pharmaceuticals.
• Delivered an end-to-end structural biology pipeline , enabling a 40% reduction in early-stage hit-to-lead discovery time 
• Hands-on developed a "cross attention model" using PyTorch with frozen ESM2's weights to map paratopes ( with CDR3 mask) and epitopes ( with antigen residue amino acids from SabDab ) , and derive binding probabilities 
• Atomic-Level 3D Docking - Deployed optimized structural prediction models (AlphaFold/BindCraft/ OpenFold/Boltz-2) on AWS g5.xlarge GPU instances to predict high-resolution 3D structures of the Her2-VHH complexes.
• Directed an engineering team of 12+ AI engineers, ML Ops specialists, and software developers. Managed direct reports , owning hiring, performance reviews, and succession planning to build a culture of continuous learning.
• Established strict Responsible AI and security frameworks for enterprise generative AI adoption. Partnered with LLM Compliance team to design zero-egress LLM architectures, implementing rigorous access controls, human-in-the-loop safeguards

DocGen — AI-Assisted Document Generation for GxP Manufacturing ( for CRDMO - Sai Life Sciences)

• DocGen , developed for a leading CRDMO, is a GMP compliance initiative . It turns the slowest step in GxP operations — authoring and revising controlled documents — into an AI-assisted workflow with human QA approval at every release. It generates new SOPs and technical packages grounded in the site's own documents and the regulatory corpus (ICH Q7/Q9/Q10/Q12, FDA guidance, EU GMP, PIC/S), and executes controlled modifications to existing documents with full traceability.

• The Project ensures coverage of 21 CFR Part 11 use cases- immutable audit trails , electronic signatures , access and authrority controls and ALCOA data integrity.
• Hybrid retrieval knowledge base - bge-m3 dense + SPLADE sparse embeddings with QDrant vector store ( for GxP compliance). Claude Sonnet 4.5 / LLaMA 3.3 70B LLMs via AWS Bedrock

RetroSynthesis AI : Optimal synthesis Pathway Platforms for Pharma Molecules ( for CRDMO - Sai Life Sciences )

• Architected a full-stack AI-powered retrosynthesis platform, reducing synthesis route identification from weeks of manual analysis to under 10 minutes. Integrated AiZynthFinder MCTS (878,000 USPTO reactions), IBM RXN & Sigma Aldrich APIs with Claude Sonnet as LLM

GxP Doc Retrieval System : AI Powered Document Intelligence System ( for Takeda Pharmaceuticals ) 

• Architected an AI-powered platform for retrieval of SDLC documents from various sources like Veeva Vault , JIRA , Confluence and sharepoint 
• Document retrieval agents were implemented with LangGraph Framework with deployments done with AWS Fargate instances 

• Architected a Product Safety Incident Classification (PSIC) platform to automatically assess severity of customer complaints across Smoke Detectors, CO Detectors, and Fire Extinguishers, reducing manual legal review effort.

• Designed the end-to-end data pipeline: scheduled jobs extracted daily customer incidents from Salesforce into Snowflake, and AWS Glue ETL jobs loaded the curated data into SQL Server on AWS to feed the ML classification model

• Used Glue crawlers and the Glue Data Catalog to maintain table metadata and schemas, making pipeline data discoverable across teams.Built data quality checks for incomplete incident records.

• Maintained product reference data and mapped incoming incidents to the master product list, so classification stayed consistent across product lines.

• Preserved Salesforce case IDs( lineage) end to end, giving traceability from each source record to its model classification.

• Partnered with legal and business leadership to define the product roadmap. Governed the engineering budget by choosing on-premises LLM deployment over SaaS APIs, controlling inference costs while keeping data private.

• STELLAW | AI-Powered Legal Research for Indian Jurisprudence

  • Architected  STELLAW, an AI-powered legal consultation and research platform purpose-built for Indian legal professionals

• CapitalMind | LLM-Powered Capital Markets Research Engine

  • Architected CapitalMind, an AI-native investment intelligence platform enabling investment bankers, equity analysts, and capital markets teams to analyze institutional behavior using natural-language queries.

InHouse CoPilot Solution

• Architected and deployed an in-house AI Copilot for enterprise .NET / C# codebases, enabling secure, context-aware developer assistance without reliance on external SaaS tools. Integrated Cline(RooCode) with .NET C# MCP Server

IPaaS Solutions ( ws02 )

• Led the enterprise integration layer feeding SAP. Built automated SFTP ingestion pipelines on WSO2 to pull invoice data from [N] external servers, with validation, transformation into SAP formats and audit logging for financial reconciliation.
• Built WSO2 Data Services to read from and write to SQL Server, exposing database operations as governed services instead of direct database access.
• Published integrations through WSO2 API Manager and set API standards for versioning, throttling and contracts, so consuming teams could integrate safely without breaking changes.

ML | Systems Analytics Lead - Capacity Planning & Forecasting 

• Owned the resource allocation and capacity strategy for 8,500+ Linux servers. Utilized ML forecasting to drive enterprise-wide infrastructure budget optimization, clearly communicating financial risks and ROI to senior banking executives.

• Ordinary Least Squares (OLS) linear regression, producing an interpretable relationship of the form y = mx + c for capacity extrapolation.

• Led multi-team delivery for a 20M SGD mobile banking re-platform, managing a diverse engineering workforce of 70 engineers across data, security and product teams, and developing technical leaders from within the IC ranks.
• Delivered the backend onto Kubernetes, running 16 containerised JBoss application servers hosting the Java services that powered the mobile banking platform.
• Owned engineering practices across the programme: TDD, BDD, DevOps and test automation.

• Responsible for Capacity Planning for new LifeStyle2.0 CARDS+ CAT1 Mobile Application Platform 
• Responsible for Application and Infra Deployment Readiness

• Forex Time Deposit , 1FA for low and medium risk transactions
• Solution Architect for Implementation of "6- tier" and "3-tier " Zero Down Time resiliency projects for HK & SG Mobile Banking

• Project manager for Cards Remodeling project 
• Responsible for deliverables on developments on ETL , SORs , Integration 

• Responsible for reverse engineering approach to re-engineer front end applications of Branch IT platform ;De-compilers , porting legacy VC++ 2.0 to .NET framework 
• IBM MQ Client Code migration to .NET framework

• Maintenance /enhancement of Repo Trading Platform- GTRepo - Fixed
  Income Prime Brokerage division 
• 24x7 L2-L3 support for  settlement systems flow of high worth , sensitive trades 
• End to end orchestration for Decommission 

• Solution Architect cum Technical Manager  for  implementing Remote IP , a noval approach to connect to telecom partner's OXO server without use of public IP
• Responsible for L3 maintenance of OXO product line

• Tech lead for Omni PCX Management Console ,Open Telephony Server 
• Ensuring adherence to SLAs on L3 maintenance defects

• Enhancement  of OmniPCX Management Console ,Open
  Telephony Server and TAPI for soft phone interface with PBX.

• Developer for OmniPCX Management Console - on Windows MFC , Win32 , FTP & TCP .Won "FIRST CENTURY" award for having fixed first 100 defects  

• Design and development of Remote Monitoring WMI Server ,
  using TCP/Win32 API & MFC .Implemented Multithreaded Threadpool