App LogoApp name
MD

Mansi Dakhale

Open to work0 years experience

AI/Machine Learning Engineer

Nagpur, Maharashtra, IndiaTarget Roles: Backend Engineer • Full-Stack Developer • Frontend Specialist

AI/Machine Learning Engineer building LLM- and vision-driven automation.

GitHub

Standing Rank

Rank Not Available

Developer Badges

No badges earned yet

Skills & Technologies

48 skills
raglangchainlanggraphagentic aiprompt engineeringmultimodal llmsloraqlorachromadbfaisspineconegroqevaluation engineeringpdf/drawing rasterizationocrmto extractionganscnnsnlpu-netvision transformersbertmonte carlo dropoutpytorchtensorflowscikit-learnhugging facedockermlflowdvcw&bfastapiflaskstreamlitceleryredisci/cdgitpytestreacttypescriptdicom/nifti processingpostgresqlmongodbnumpypandaspythonsql

Work Experience

AI Engineer Intern

Evomaton

Jun 2026 - Present Nagpur, Maharashtra, India
  • Building an LLM-driven pipeline to extract Material Take-Offs (MTO) from complex construction drawings in EPC domain
  • Using a multimodal LLM to extract dimensions, openings, and material specs, with a Python layer for quantity calculations; completed feasibility analysis and technical documentation, alongside AI feature work on Databricks and Azure

AI Engineer Intern

Blu Parrot

Feb 2026 - Mar 2026 Nagpur, Maharashtra, India
  • Built an AI-powered agri-advisory platform for multi-region crop recommendations across Indian states, combining OCR-based document ingestion with rule-based agronomic logic and ML models for soil (NPK) and climate data

Projects

OpsPilot AI

View Project
FastAPILangGraphChromaDBReactGroqAWSPostgreSQLRedisCeleryDocker
  • Architected a multi-agent platform (LangGraph + Groq LLaMA 3.3 70B) with a Planner agent that routes queries to specialist nodes for RAG document Q&A, tool execution, code generation, and report writing
  • Built a RAG pipeline with ChromaDB and MMR retrieval plus a two-pass fact-verification layer; implemented asynchronous PDF ingestion via Celery and Redis with OCR fallback
  • Deployed end-to-end on AWS (EC2, RDS, S3) with a React 19 streaming frontend, containerized using Docker Compose

AI Product Creative Generation Workflow

View Project
FastAPILangGraphGroqCeleryRedisReact
  • Automated end-to-end ecommerce creative generation (5 images, 2 videos, campaign strategy) using a 6-stage multi-agent LangGraph workflow
  • Built FastAPI APIs with asynchronous job execution via Celery and Redis, supporting CSV-based bulk processing
  • Deployed the full workflow on Render and Vercel for production use

Agentic RAG Research Assistant

View Project
LangChainChromaDBStreamlit
  • Built a hierarchical multi-agent RAG system reasoning across 50+ research papers
  • Implemented an LLM-as-a-judge (RAGAS) evaluation pipeline scoring faithfulness, relevancy, and context precision/recall
  • Achieved 92.4% retrieval precision@5 and 84.7% keyword accuracy by tuning chunking and embeddings; shipped a Streamlit app with fully offline model serving via Ollama/llama.cpp

UGSA-TSGAN: Tumor-Attentive GAN for CE-MRI Synthesis (M.Sc. Thesis)

View Project
PyTorchGANsMedical Imaging
  • Curated a specialized dataset of 239 patient records from the ISPY1 cohort and wrote custom PyTorch training loops from scratch
  • Designed a novel Uncertainty-Guided Soft-Attention Block (UG-SAB) fusing feature maps with Monte Carlo Dropout uncertainty maps, cutting annotation needs by 40% via semi-supervised pseudo-labelling
  • Beat the fully-supervised baseline by +6.5% Global SSIM(p=0.0072) and +9.0% Tumour ROI SSIM(p=0.0029); cut MAE by 47% (0.086 to 0.0454); synthetic CE-MRI retained 69.7% diagnostic utility

Leaderboard Standings

Leaderboard Position Pending

Global test scores, peer standing percentiles, and algorithm leaderboard ranks are updated dynamically.

Assessment Highlights

Assessments Not Completed

Coding evaluations, system assessment results, and conceptual score badges will appear here after taking a test.

AI Collaboration Score

AI Collaboration Score Pending

Developer coding behavior, assistant cooperation, and AI pair-programming indicators are evaluated during live coding sessions.

Role Compatibility Profile

Role Compatibility Analysis Pending

Custom matching reports, candidate role compatibility percentiles, and core engineer strength profiles are processed once conceptual code screenings are complete.

Achievements

IIT JAM 2024 Rank

IIT JAM 2024: All India Rank 320 in Statistics

IIT Guwahati Summer Project 2025

IIT Guwahati Summer Project 2025: Inventory Optimization using Advanced SQL Analytics

About Details

Professional Bio

AI/Machine Learning Engineer building LLM- and vision-driven automation for engineering and EPC workflows. Experienced in designing multi-agent workflows with LangGraph and shipping low-latency inference APIs with FastAPI on AWS/Azure.

M.Sc. in Artificial Intelligence & Machine Learning in Artificial Intelligence & Machine Learning

Indian Institute of Information Technology Lucknow (2024 - 2026)

B.Sc. in Statistics, Computer Science & Mathematics in Statistics, Computer Science & Mathematics

RTMNU (2018 - 2021)

Languages: English, Hindi