This website uses cookies

Read our Privacy policy and Terms of use for more information.

Marshall Deutz of the Product & Solutions team at SpotDraft, which sponsored the session, moderated this How to Contract webinar with Aprajita Tyagi, Senior Corporate Counsel at Miro, and Anya Ryjkova, SVP, Legal at BNY. Aprajita leads partnership transactions and supports large strategic customer deals, and she has stepped more and more into a legal engineering role, building her own AI agents for the team. Anya trained as a computer scientist and mathematician before law, spent a decade at Silicon Valley tech companies, and now works at a heavily regulated financial institution. Together they brought the deal-level view and the data-structure view, and Marshall added what he sees from the vendor side when legal teams adopt new tools.

Want unlimited access to 110+ webinar replays? Our paid members get permanent access to webinar recordings from 2024 and 2025, plus 30+ hours of courses courses, certifications, and our massive expert library. If you just want an individual webinar, register in advance to get access to the recording for 14 days. You can sign up for our auto registration list so you are registered for every webinar we host. Learn more about our membership →

The conversation moved from what actually sits inside our contracts to why portfolios are still so hard to query, how extracted data becomes intelligence, and what our negotiation history says about our playbooks. The speakers also covered how other teams can use contract data, what makes the data trustworthy, who should own its quality, and where to start this quarter.

Here are our top ten takeaways from the speakers' comments during the webinar:

  1. Look at how provisions work together. Aprajita said the obvious data points like pricing, liability caps, renewals, and AI training rights only tell part of the story. A business team might ask whether referral partner agreements include termination for convenience, and a quick Control + F answers that. Another provision in the partner program could still give those partners deal protection for six months after termination. The relationship between provisions, and the patterns across the whole portfolio, give the business a much better answer.

  2. Get your contracts into one place first. Marshall called it operational readiness, and both speakers came back to it at the end. Anya said even the fastest AI tool gets confused when the underlying dataset is missing the latest amendments. Contracts scattered across email, personal folders, and SharePoint sites, or dumped into a CLM with poor metadata, won't give you reliable answers. Centralize the executed agreements with every amendment, order form, and exhibit that goes with them.

  3. Aim for data you can act on. Aprajita explained that searchable data lets you find a reference, like whether a contract auto-renews. Usable data answers the business question behind it, like when a renewal notice has to go out, who receives it, and who inside the company tracks it. She remembered sending post-acquisition assignment notices by searching hundreds of contracts for the word assignment and reading each provision by hand. Today she can extract that information in a form she can act on right away.

  4. Start with a real business question. Tools that can extract everything don't mean we should extract everything. Aprajita recommended picking a real problem, like missed renewal deadlines, and pulling the relevant data from a defined set of recent agreements. From there you can build a skill that pings the right person when a notice is due. Anya added that deciding how to structure the data deserves careful thought up front, not a last-minute scramble.

  5. Let your negotiation history shape your playbook. Aprajita compares recently negotiated contracts against the standard template and asks AI to list and categorize the deviations. The patterns show what we regularly concede. When every customer agreement covering AI features ends up saying "we will not train on your data," it's time to add that language to the template. The same history shows which regulatory terms you've accepted for customers in industries like telecom or medical.

  6. Record the outliers and why you agreed to them. Anya watched teams uncap liability at quarter end because they were desperate to close. Those deals land in the core dataset and confuse the AI tools that rely on it. Escalate the big deviations and note in the metadata that management approved a risky call. The tool then treats the deal as an outlier, and your contract manuals and training for newer lawyers reflect the real baseline.

  7. Check AI's work against the source. Aprajita warned that AI tools hallucinate and sometimes create information just to answer the question. Ask the tool to cite the source, quote the exact provision, or link you to it, and keep a human in the loop for high-risk items. Anya asks for section numbers so she can verify the output and tells the tool which agreements belong to the same package, especially after a company name change. Once a query works, save it as a template your colleagues can reuse and improve.

  8. Share what the data shows beyond legal. Contracts hold business and operational information, not just legal terms. Aprajita described go-to-market teams measuring the impact of an expanded reseller territory, procurement spotting vendor concentration and forgotten auto-renewals, and InfoSec comparing what we promised customers with what we get from vendors. Anya uses workload data to show leadership what a sales push will demand from legal. Marshall said that once his sales colleagues saw how long negotiations really took, they planned their pipeline differently.

  9. Improve older contracts at renewal. Anya deals with decades-old contracts missing data protection addenda, audit clauses, or current terms. Her advice was to leave a working revenue relationship alone. Tag the contract for renewal, prepare the language you want to add, and fill the gaps when the renewal comes around instead of papering a brand new contract. Then capture data properly going forward.

  10. Give data quality a named owner. Anya said one person needs their name on contract data quality, with performance reviews and incentives tied to it. Users need a feedback loop to tell that person what isn't working, and leadership needs to back the work with resources. Marshall saw the opposite from the vendor side, where one person picks a tool with no executive sponsor and it turns into shelfware once they leave. Clear ownership keeps the system improving.

Subscribe to Stay in the Loop

Our weekly newsletter keeps you current on upcoming How to Contract webinars and brings you recaps like this one when you can't make the live session. Subscribe now and get the practical takeaways delivered to you.