Senior Full-Stack Engineer · AI/ML · India
Anshuman Nayak
I build production systems (payments, data pipelines and LLM-powered products) and publish research applying machine learning to soil science.
- 6+ yrs
- shipping production software
- 10B+
- records migrated monthly via ETL
- 1M+
- customer queries handled by LLM chatbots
- 12
- peer-reviewed papers, 161 citations
About
I'm a full-stack engineer with six years of experience taking systems from design to production: a payment gateway handling 15,000+ transactions a day, ETL pipelines moving 10 billion records a month, and a low-code API platform that cut delivery time from 14 days to 2 hours for 2,000+ data scientists. More recently I've worked on LLM training data and domain-specific AI evaluation with Micro1 AI and Scale AI.
I hold an M.Tech from IIT Kharagpur, and alongside industry work I collaborate on research in precision agriculture: smartphone-based soil sensing, proximal and remote sensor fusion, and generative-AI-assisted soil carbon estimation. That work has produced 12 peer-reviewed papers and a patented smartphone imaging technology.
I'm open to senior engineering and applied-ML roles.
Skills
- Languages
-
- Python
- JavaScript
- TypeScript
- Go
- Rust
- C
- AI / ML
-
- GPT
- Claude
- Gemini
- LangGraph
- LangChain
- RAG
- Multi-agent systems
- AI evaluation
- TensorFlow
- OpenCV
- pandas
- NumPy
- Frameworks
-
- Node.js
- Express
- NestJS
- React
- Next.js
- Ionic
- Django
- Flask
- FastAPI
- GraphQL
- REST
- Data, cloud & tools
-
- PostgreSQL
- MySQL
- MongoDB
- Firebase
- Snowflake
- AWS (EC2, EKS, RDS, S3, Amplify)
- Docker
- Terraform
- Nginx
- Git
- Jira
Experience
-
Aug 2025 – Present
Human Data Manager
Micro1 AI · Consultancy
- Led a cross-functional team of 53 subject-matter experts across data science, healthcare, physics and executive leadership.
- Curated and validated full-stack code repositories (Node.js, React, Python, Go, Ruby) for LLM training datasets.
- Built data ingestion pipelines that prioritise human-generated over synthetic content to improve model reasoning.
- Set up multi-stage quality review and human-in-the-loop workflows to raise alignment and code fidelity.
LLM training data · HITL · Team leadership · Code review
-
Jan 2025 – Aug 2025
Engineering Expert, Human Frontier Collective Fellow
Scale AI · Fellowship
- Evaluated AI model solutions in agricultural and environmental engineering, identifying errors, limitations and failure modes.
- Wrote domain-specific test questions and benchmarks to assess model reliability in agricultural technology.
- Supplied expert corrections for model mistakes, feeding improved training data and evaluation datasets.
- Guided model behaviour in soil science, precision farming and environmental monitoring.
AI evaluation · Benchmarks · Soil science · Precision farming
-
Feb 2024 – Jun 2026
Software Engineer II
Zeno Health · Remote
- Engineered a payment gateway processing 15,000+ transactions a day with 99.9% uptime and sub-100 ms latency.
- Architected ETL pipelines migrating 10 billion+ records every month with zero data loss.
- Deployed GPT-4 chatbots handling 1M+ customer inquiries a year, cutting manual support volume by 60%.
- Orchestrated multi-agent systems (GPT-4, Gemini, Claude) that automate product-development workflows, improving team efficiency by 70%.
- Built Zeno Rewards, serving 250+ offline stores and 2 million monthly active customers.
- Led and mentored 10 developers, set code-quality standards and introduced self-healing AI agents for automated code review.
React · Node.js · Python · MySQL · GPT-4 · Gemini · Claude
-
Jul 2022 – Feb 2024
Software Engineer
Carelon Global Solutions · Gurugram
- Created a low-code platform for the Data API Gateway, reducing development time from 14 days to 2 hours.
- Built its fault-tolerant core, supporting 2,000+ data scientists and 5+ production applications at 99.9% uptime.
- Owned the data-model lifecycle, from schema design to pipeline optimisation, with real-time alerting and incident response.
Low-code platforms · API gateway · Observability
-
Jan 2019 – Mar 2022
Full-Stack Developer
Nex Verse Pvt. Ltd.
- Built the interface for a hybrid social app that grew to 8,000+ users and 15 businesses.
- Developed the Node.js and MongoDB backend, added AWS EC2 autoscaling and cut video-processing latency by 40%.
- Shipped push notifications, in-app messaging, A/B testing and analytics with Firebase.
Cordova · Node.js · MongoDB · AWS EC2 · Firebase
Education
-
2020 – 2022
M.Tech, Agricultural Systems & Management
Indian Institute of Technology (IIT) Kharagpur
CGPA 9.51 / 10 · GATE 2020 All-India Rank 67
-
2016 – 2020
B.Tech, Agricultural Engineering
Odisha University of Agriculture and Technology (OUAT), Bhubaneswar
CGPA 8.50 / 10
Honours
- Patent granted for smartphone-based detection of nitrate and phosphate in soil and water samples
- Recognised Scholar in Digital Agricultural Technology, Geological Society of America, Kyoto, Japan
- All-India Rank 67, GATE 2020 (Agricultural Engineering)
Certifications
- GraphQL Developer Associate, Apollo
- Meta Back-End Developer Program, Meta
- Data Analytics with Python, NPTEL (Silver Elite Award)
- Responsible and Safe AI, NPTEL
- JavaScript Algorithms and Data Structures, freeCodeCamp
Projects
-
Haiti Digital Soil Mapping · SolumSense
Country-wide digital soil-mapping platform with nutrient analysis and precision fertilizer recommendations, used by policymakers, fertilizer companies and smallholder farmers.
Next.js · FastAPI · MySQL
View project → -
Fusion CMS
Generates GraphQL and REST backends from SQL and NoSQL sources with no hand-written code. Deploys as a monolith or microservice with custom middleware.
TypeScript · Next.js · GraphQL · REST
View on GitHub → -
SpectraML
Open-source AutoML platform for spectroscopic data: preprocessing, feature engineering, model selection, PCA/t-SNE visualisation and evaluation, with reproducible Jupyter pipelines.
Django · Python · Plotly · D3.js · Jupyter
-
Mobile Plant Disease Detector
Offline-capable app that lets farmers identify plant diseases from a smartphone photo, running deep-learning models on the device.
TensorFlow.js · OpenCV.js · WebAssembly · Ionic
Read the paper → -
Spectagro
Agri-tech platform for soil testing and analytics, nutrient recommendations, crop-area estimation and sourcing.
Agri AI · Analytics
View project → -
Zeno Rewards
Loyalty and offers product for 2 million monthly customers across 250+ pharmacy stores.
React · Node.js · MySQL
View project → -
Agriculture Tools
Free field calculators: fertilizer dose, seed rate, spray mix, rainfall analysis and machinery hiring cost.
Calculators · Open access
Open tools →
Research
15 publications 161 citations h-index 6 i10-index 4
- 2026
- 2026
- 2026
- 2026
- 2026
- 2026
- 2025
- 2025
- 2025
- 2025
- 2024
- 2024
- 2023
- 2023
- 2023
Citation counts from Google Scholar, October 2026 · ORCID · Citation lookup tool
Contact
The fastest way to reach me is email. I usually reply within a couple of days.
- Emailanshuman_n@outlook.com
- LinkedInanshuman-nayak
- GitHubanshumanfs
- LocationIndia