Distribution · Partner networks
A sales and support platform for a distributed partner network
Partners ask who on their team is inactive and what blocks the next level, and get numbers for their own part of the network only. Access is enforced at the SQL level.
- Status
- In production
- Timeline
- about 12 months: two phases, about 8 of them in active development
- Stack
- Python / FastAPI
- OpenAI API (Assistants, Whisper)
- spaCy
- MS SQL Server
- Neo4j
- React / TypeScript
The task
A company running a distributed partner network needed a sales and support platform. The first job: partners and team leads could not quickly see who was inactive or what blocked the next level. The data sat in a back-office database and several internal APIs. Direct LLM access to the database was off the table: personal data, other teams and the schema had to stay closed.
What we built
- We built a chat and voice assistant: each question is routed to the client’s internal APIs or to model-generated, read-only SQL against the back office, and answered in business terms.
- The model does not guess the schema: table descriptions and a relationship graph point it to the right tables and joins, and every generated query is logged for review.
- No query reaches the database unchecked: read-only access, a table whitelist, mandatory per-user scoping, blocked personal-data columns, response sanitisation and rate limits.
- On top of the database sits a network-analytics layer: team activity, sales-volume gaps, level blockers and reward forecasts. Admins get a panel with chat statistics, logs and promotions.
- The platform design also covers buyer and marketing modules: personal recommendations, repeat purchases, loyalty tiers, segmentation, and campaigns with A/B testing and push and email delivery.
Outcome
- Their own slice
- a partner gets figures for their own team only: a filter on the current user is added to every SQL query
- 6 layers
- of data protection between the language model and the database, from blocked personal-data columns in SQL to rate limits; counted in the code
- Accepted on live data
- acceptance checks ran on the client’s working back-office database. We do not give the acceptance figures here: the client has not confirmed consent to publish them, and they cannot currently be reproduced from the code on our server
Services
More cases
Aviation · aircraft components trading and MRO
PilotAircraft parts procurement: from email to a managed pipeline
An AI agent reads incoming RFQs, compares supplier quotes and checks trace documents. A separate deal-tracking system moves each deal through mandatory stages.
- Full loop
- the AI loop ran end to end on a live procurement process, from an incoming RFQ to a team notification. This was a pilot and the figures are our own so far: the client has not confirmed them yet
- 3
- types of trace documents checked by the agent: FAA 8130-3, EASA Form 1 and COC. Like the figure above, it still awaits the client’s confirmation
IT consulting · Lead qualification
In-house productA pilot plan for the visitor, a qualified lead for sales
We replaced the contact form with an AI consultant: it interviews the visitor, returns a pilot plan with KPIs and risks, and sales gets a lead with budget and timeline.
- 1 day
- to build a working MVP: frontend, backend and deploy configs
- 16 days
- from first commit to production with HTTPS and an issued certificate on its own domain
Marketplace e-commerce · Sales analytics
In productionProfit per SKU across three marketplaces, checked against manual figures
Five seller accounts on WB, Ozon and Yandex Market merged into one SKU-level P&L. Variance from the client’s manual figures on a control SKU for one month: 0.00%.
- 0.00%
- variance vs the client’s manual benchmark across units, revenue, COGS, profit and margin for a control SKU over a month; the threshold was 1%
- 1.18M rows
- of sales across 5 accounts and 12 months processed; 10.7% were ordered and settled in different months
Estimate the payback of a similar process
AVA lets you work out the economics of your process on your own figures, with no form and no price from us. It shows what the manual work costs, what cash effect automation would give and what budget would pay back. If the effect does not even cover running costs, it says so.
Estimate the paybackDiscuss a similar problem
Describe the process and the data you have. Before any work starts we will tell you what can be measured in money, what cannot, and whether it is worth doing.