We build production AI systems for communities that need them most.
We need experience to get a job. We need a job to get experience. Let's break the loop together.
So "years of experience" doesn't start at zero.
From student to engineer. For real.
If you want hands-on engineering tied to partner needs, there is room to contribute here.
Faculty & Advisors
“Most agentic AI research focuses on what these systems can do. We at AnacodicAI Labs focus on who they're for...”
“Research rigor means being able to show your work at every step — the methods, the assumptions, the limitations...”
Why we exist
Students need a hire to prove what they can do. Partners need a budget line for tools that would matter—when both gates stay shut, nothing moves.
We build in that opening: with the people waiting on the other side, in public, under the same evidence bar we use in peer-reviewed work.
44M teachers short by 2030. 66% of Malawi schools: 90+ students per teacher.
We build: Terrier Grader
Source · OECD 2024 →
Rural clinics with no access to latest research. < 1 psychiatrist per 100K in low-income countries.
We build: ClinicalSearch · Mindwell
Source · PMC 2025 → · WHO 2024 →
Skill-rich artisan communities locked out of premium markets.
We research: Weave Forward
Source · Field research 2024 → · Challakere, Punjab →
How we bridge it
You bring the skills; they hold the mandate. We run the collaboration like research and ship like production—clear scope, honest review, systems that still work after the semester ends.
What you get
Students leave with work they can defend in a room. Partners leave with systems they could not have sourced from a catalog—or an RFP—alone.
Where we build
Rooted at Boston University.
Engineered to hold where infrastructure, capital, and attention run out first.
Faculty-advised research. Two published in peer-reviewed journals. Two open for student collaborators.
AI Energy Research
Empirical benchmarking of energy consumption and carbon emissions across frontier LLMs on 5 MMLU domains. Establishes a per-query carbon cost framework with a 4.3× energy differential between model endpoints and 12.1% cross-domain bias in energy allocation. Foundation for the CCI service integrated into the ClinicalSearch pipeline.
Rashanjot Kaur, Triparna Kundu, Kathleen Marshall Park, Eugene Pinsky
Optimization AI
Multi-agentic framework for supply chain optimization under carbon regulatory constraints. Introduces the CASP metric a weighted harmonic mean for evaluating operational resilience across energy transition scenarios. Models socio-technical dependencies that static optimization ignores.
Rashanjot Kaur, Triparna Kundu, Bhanu Sharma, Kathleen Marshall Park, Eugene Pinsky
Faculty-advised research — team forming
→ Core contributors included as co-authors
Healthcare AI
Multi-agent architecture for mental health support with PII boundary enforcement at the system level, not the application level. Ephemeral session design prevents cross-interaction data leakage. Structured agent delegation across intake, assessment, resource matching, and escalation under HIPAA regulatory constraints.
Apply by April 30, 2026
Faculty-advised research — team forming
→ Core contributors included as co-authors
Evidence Retrieval AI
Namespace-partitioned multi-agent retrieval architecture with source-grounded synthesis for clinical evidence access in settings without research librarians or database subscriptions. Nine domain-specialized retrieval agents coordinated through orchestrator-driven routing on AWS Bedrock. Citation provenance preserved end-to-end to prevent hallucinated evidence.
Apply by April 30, 2026
What we're building right now
Academic AI
Production evaluation pipeline with rubric-aligned generative assessment and self-consistency verification. Retrieval-augmented feedback synthesis across configurable assessment criteria. 500+ documents per evaluation cycle, 3 institutional deployments, 60% overhead reduction.
RAG Systems · Prompt Engineering · NLP
Apply by April 30, 2026
Healthcare AI
Privacy-preserving multi-agent system with ephemeral session architecture and PII boundary enforcement. Structured agent delegation across intake, assessment, resource matching, and escalation pathways.
Multi-Agent Systems · LangGraph · Healthcare AI
Apply by April 30, 2026
Evidence Retrieval AI
Multi-agent evidence retrieval system with 9 domain-specialized agents coordinated through orchestrator-driven routing on AWS Bedrock. Parallel consultation across namespace-partitioned vector stores with source-grounded synthesis.
Multi-Agent RAG · Pinecone · Medical NLP
Apply by April 30, 2026
Business AI
5-agent orchestration layer on AWS Bedrock with Model Context Protocol integration and dynamic tool selection. Function-calling coordination across domain-specialized agents. Deployed across 4 geographic regions at 90% cost reduction.
AWS Bedrock · MCP Protocol · Agent Orchestration
Apply by April 30, 2026
Write code that someone is waiting for.
Founded at BU · Systems backed by research · Volunteer-run
Collaborating with: BU · Cleveland Clinic · Bharattap
