AI assistant for biostimulant R&D workflows

Accelerating biostimulant R&D with an AI-powered research assistant, targeting a 50% reduction in manual workload

Scroll down to read more

Project highlights

Industry: Agricultural Biotechnology

Client services: AI Business Transformation

Start date: December 2025

Location: United States

Team size: 6 members

Duration: 7 months

About the client

Our client is an agricultural biotechnology company transforming organic waste into sustainable biostimulant products that help crops withstand environmental stress.

By exploring the potential of materials such as algae and biosolids, the company develops innovative formulations designed to improve crop resilience against drought, heat, and other extreme growing conditions.

Their mission is to turn underutilized biological resources into valuable agricultural solutions that support sustainability, food security, and human health.

Business challenges

The client’s R&D team relied on scientific literature, spreadsheets, presentations, and research reports scattered across disconnected sources. Finding relevant information, identifying relationships, and evaluating potential formulations required significant manual effort.

This fragmented approach created several challenges:

Time-consuming research: Scientists spent excessive time locating, reviewing, and consolidating information instead of advancing research.
Disconnected knowledge: Valuable scientific findings were scattered across documents, making them difficult to access, compare, and reuse.
Limited research capacity: Manual workflows restricted the team’s ability to explore new ingredients and test formulation hypotheses.
Lack of relationship visibility: Identifying connections between materials, metabolites, microorganisms, and biological effects required extensive manual analysis.

The company needed a more efficient, scalable way to transform its accumulated research into actionable scientific knowledge.

Goals set to Achievion

Achievion was engaged to develop an AI-powered research environment that would streamline scientific knowledge discovery and accelerate the identification of promising biostimulant ingredients and formulations.

The primary objectives included:

Centralize research knowledge by transforming fragmented scientific literature into a structured, searchable knowledge base.
Reduce manual R&D effort by automating information extraction, document analysis, and knowledge retrieval.
Enable intelligent research exploration through natural-language questions and AI-generated, evidence-backed responses.
Reveal scientific relationships between materials, extraction methods, biological effects, and potential formulations.
Accelerate formulation discovery by helping researchers identify promising ingredients and evaluate new hypotheses more efficiently.

Solution

Achievion developed an AI-powered research assistant that transforms fragmented scientific literature into a connected, searchable knowledge base.

The platform uses natural language processing (NLP), AI-driven information extraction, and retrieval-augmented generation (RAG) to process scientific documents, identify relevant entities, and uncover relationships between materials, metabolites, microorganisms, and biological effects.

Instead of manually reviewing disconnected documents, researchers can now explore structured scientific knowledge, search for relevant materials, compare findings, and ask questions through an intuitive conversational interface.

The solution combines several core capabilities:

Automated knowledge extraction: Processes scientific literature and extracts structured information about materials, biological effects, metabolites, and microorganisms.
Connected knowledge base: Organizes extracted information into a centralized repository that makes scientific relationships easier to discover.
Semantic search and filtering: Enables researchers to locate relevant findings using keywords, scientific concepts, and entity-based filters.
AI-powered research assistant: Uses RAG to generate contextual answers grounded in the client’s scientific knowledge base.
Source traceability: Provides citations linking AI-generated responses to the underlying research documents.
Researcher-friendly web interface: Brings document discovery, knowledge exploration, and conversational AI together in one application.

The resulting platform replaces fragmented research workflows with a more efficient, evidence-driven approach to scientific exploration.

Key features of the product in detail

Intelligent Document Search

Researchers can quickly locate relevant scientific publications, ingredients, and formulation data through an intuitive search interface.

The system identifies matching materials and associated scientific information, reducing the time spent navigating individual documents and making research findings easier to discover and reuse.

Advanced Knowledge Filtering

Entity-based filters allow researchers to narrow results by materials, plants, biological effects, and other scientific attributes.

This structured approach simplifies the exploration of complex relationships and helps researchers identify relevant information without manually reviewing large volumes of literature.

AI-Powered Q&A Assistant

A conversational AI assistant enables researchers to explore scientific knowledge using natural language.

Users can ask questions, investigate relationships, and request additional details through an interactive dialogue. Powered by RAG, the assistant retrieves relevant scientific evidence to support its responses, making research exploration faster and more intuitive.

Evidence-Based Responses with Citations

Every AI-generated response is supported by references to relevant source documents.

Researchers can trace findings back to the original literature, verify supporting evidence, and incorporate insights into their research with greater confidence.

This transparency helps maintain scientific rigor while making AI-assisted research more practical and reliable.

Business outcome

Achievion delivered an AI-powered research platform that transforms how the client’s scientists access, explore, and apply scientific knowledge.

By replacing fragmented research processes with a centralized, evidence-backed AI assistant, the solution creates opportunities to significantly reduce manual effort and accelerate biostimulant R&D.

Key business benefits include:

Up to 50% projected reduction in manual R&D workload: Automated knowledge extraction, search, and AI-assisted exploration are expected to substantially reduce time spent on repetitive research activities.
Faster ingredient discovery: Researchers can identify and evaluate promising biostimulant ingredients without manually searching multiple disconnected sources.
Accelerated formulation research: Structured scientific knowledge makes it easier to shortlist promising hypotheses and explore potential ingredient combinations.
Improved research efficiency: Centralized information and intelligent retrieval reduce repetitive document reviews and manual information tracking.
More confident, evidence-based decisions: Traceable scientific references help researchers validate findings and make better-informed R&D decisions.
Scalable foundation for future innovation: The reusable RAG architecture provides a foundation for further optimization, expanded research capabilities, and future predictive modeling.

The solution establishes a new foundation for AI-assisted scientific discovery, enabling the client to spend less time searching for information and more time developing innovative, sustainable agricultural products.

Timeline

2 Months

Design and Proof of Concept Development

Achievion worked closely with the client’s research team to validate the technical approach and establish the foundation for AI-powered knowledge extraction.

Key activities included:

Conducting discovery sessions to understand R&D workflows and research requirements.
Reviewing scientific source materials and validating the proposed data schema.
Developing and testing initial information extraction logic.
Evaluating the feasibility of structuring and connecting scientific knowledge.

5 Months

MVP Development

Following successful concept validation, Achievion developed and stabilized the AI-powered research platform.

Key activities included:

Building the document ingestion and processing pipeline.
Implementing structured metadata storage and semantic search capabilities.
Integrating the RAG-powered conversational assistant.
Developing the researcher-facing web application.
Testing, refining, and stabilizing the platform for practical R&D workflows.

Team

Product Manager

Solutions Architect

AI/ML Engineer

Full-Stack Developer

MLOps Engineer

QA Engineer

Tech Stack

AI

Amazon Bedrock

Amazon AgentCore

Amazon SageMaker Embeddings

Application Layer

Python

FastAPI

Docker

React

AWS Amplify

Data Layer

Amazon Aurora PostgreSQL

Amazon OpenSearch Service

Amazon S3

Infrastructure

AWS Fargate

AWS Lambda

Amazon API Gateway

AWS IAM

Amazon CloudWatch

AWS CodePipeline

AWS CodeBuild

Get in touch to move your AI transformation
or product idea from concept to execution.